Systems, methods, and programs

Abstract

To provide a navigation function for an autonomous vehicle using a camera.SOLUTION: A system for navigating a host vehicle may include at least one processing device. The processing device may also be programmed to receive, from an image capturing device, an image representing an environment of the host vehicle, analyze the image, and identify an object in the environment of the host vehicle. Further, the processing device may also be programmed to determine a location of the host vehicle and receive map information, which is map information associated with the determined location of the host vehicle and includes elevation information associated with the environment of the host vehicle. Furthermore, the processing device may be programmed to determine a distance from the host vehicle to the object on the basis of the elevation information and determine a navigational action for the host vehicle on the basis of the determined distance.SELECTED DRAWING: Figure 41

Description

[Technical Field]

[0001] [Cross-reference of related applications] This application claims priority to U.S. Provisional Patent Application No. 62 / 652,029 and U.S. Provisional Patent Application No. 62 / 652,039, filed on April 3, 2018. All of the above applications are incorporated herein by reference in their entirety. [Background technology]

[0002] This disclosure generally relates to navigation for autonomous vehicles. [Background information]

[0003] As technology continues to advance, the goal of fully autonomous vehicles capable of navigating on roadways has emerged. Autonomous vehicles may need to consider various factors and make appropriate decisions based on those factors in order to reach their target destination safely and accurately. For example, autonomous vehicles may need to process and interpret visual information (e.g., information captured from cameras) and may also use information obtained from other sources (e.g., from GPS devices, speed sensors, accelerometers, suspension sensors, etc.). At the same time, in order to navigate to their destination, autonomous vehicles may need to identify their position within a specific roadway (e.g., a specific lane on a multi-lane road), navigate with other vehicles, avoid obstacles and pedestrians, observe traffic signals and signs, and navigate from one road to another at appropriate intersections or interchanges. The use and interpretation of the vast amount of information collected by the autonomous vehicle while it is traveling to its destination presents many design challenges. The complete amount of data that autonomous vehicles may need to analyze, access, and / or store (e.g., captured image data, map data, GPS data, sensor data, etc.) presents a challenge that could actually limit, or even worsen, autonomous navigation. Furthermore, if autonomous vehicles rely on conventional mapping techniques for navigation, the complete amount of data required to store and update maps presents an extremely difficult challenge. [Overview of the project]

[0004] Embodiments not inconsistent with this disclosure provide systems and methods for navigating autonomous vehicles. The disclosed embodiments may provide navigation functionality for autonomous vehicles using cameras. For example, not inconsistent with the disclosed embodiments, the disclosed system may include one, two, or more cameras for monitoring the vehicle's environment. The disclosed system may provide navigation responses based, for example, on the analysis of images captured by one or more of the cameras.

[0005] A system for navigating a host vehicle is provided, without being inconsistent with the disclosed embodiments. The system may comprise at least one processing device. At least one processing device may be programmed to receive at least one image representing the environment of the host vehicle from an image acquisition device. At least one processing device may be further programmed to analyze at least one image to identify objects in the environment of the host vehicle. At least one processing device may be further programmed to determine the position of the host vehicle. At least one processing device may be further programmed to receive map information, which includes altitude information associated with the environment of the host vehicle, and which is associated with the determined position of the host vehicle. At least one processing device may be further programmed to determine the distance from the host vehicle to an object based on at least the altitude information. At least one processing device may be further programmed to determine a navigation action for the host vehicle based on the determined distance.

[0006] A method for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The method may include the step of receiving at least one image representing the environment of the host vehicle from an image acquisition device. The method may further include the step of analyzing at least one image to identify objects in the environment of the host vehicle. The method may further include the step of determining the location of the host vehicle. The method may further include the step of receiving map information, which is map information associated with the determined location of the host vehicle and includes altitude information associated with the environment of the host vehicle. The method may further include the step of determining the distance from the host vehicle to an object, based at least on the altitude information. The method may further include the step of determining a navigation action for the host vehicle based on the determined distance.

[0007] A system for navigating a host vehicle is provided, without being inconsistent with the disclosed embodiments. The system may comprise at least one processing device. At least one processing device may be programmed to receive at least one image representing the environment of the host vehicle from an image acquisition device. At least one processing device may be further programmed to analyze at least one image to identify objects in the environment of the host vehicle. At least one processing device may be further programmed to determine the location of the host vehicle. At least one processing device may be further programmed to receive map information associated with the determined location of the host vehicle, which includes lane width information associated with roads in the environment of the host vehicle. At least one processing device may be further programmed to determine the distance from the host vehicle to an object based on at least the lane width information. At least one processing device may be further programmed to determine a navigation action for the host vehicle based on the determined distance.

[0008] A method for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The method may include the step of receiving at least one image representing the environment of the host vehicle from an image acquisition device. The method may further include the step of analyzing at least one image to identify objects in the environment of the host vehicle. The method may further include the step of determining the location of the host vehicle. The method may further include the step of receiving map information associated with the determined location of the host vehicle, which includes lane width information associated with roads in the environment of the host vehicle. The method may further include the step of determining the distance from the host vehicle to an object based at least on the lane width information. The method may further include the step of determining a navigation action for the host vehicle based on the determined distance.

[0009] A system for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The system may comprise at least one processing device. At least one processing device may be programmed to receive at least one image representing the environment of the host vehicle from an image acquisition device. At least one processing device may be further programmed to analyze at least one of a plurality of images to identify a first object in the vehicle environment, wherein the first object and the road on which the first object is located are at least partially obscured by a second object in the vehicle environment. At least one processing device may be further programmed to determine scale change information for the first object based on at least two of the plurality of images. At least one processing device may be further programmed to determine the lane position of the first object relative to the road lane on which the first object is located, based on the determined scale change information for the first object. At least one processing device may be further programmed to determine a navigation action for the host vehicle based on the determined lane position of the first object.

[0010] A method for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The method may include the step of receiving a plurality of images representing the environment of the host vehicle from an image acquisition device. The method may further include the step of analyzing at least one of the plurality of images to identify a first object in the vehicle environment, wherein the first object and the road on which the first object is located are at least partially obscured by a second object in the vehicle environment. The method may further include the step of determining scale change information for the first object based on at least two of the plurality of images. The method may further include the step of determining the lane position of the first object relative to the lanes of the road on which the first object is located, based on the determined scale change information for the first object. The method may further include the step of determining a navigation action for the host vehicle based on the determined lane position of the first object.

[0011] A system for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The system may comprise at least one processing device. At least one processing device may be programmed to receive a plurality of images representing the host vehicle's environment from an image acquisition device. At least one processing device may be further programmed to analyze at least one of the plurality of images to identify a target vehicle in the host vehicle's environment. At least one processing device may be further programmed to receive map information associated with the host vehicle's environment. At least one processing device may be further programmed to determine the trajectory of the target vehicle over a period of time based on the analysis of the plurality of images. At least one processing device may be further programmed to determine the position of the target vehicle relative to roads in the host vehicle's environment based on the determined trajectory and map information of the target vehicle. At least one processing device may be further programmed to determine navigation actions for the host vehicle based on the determined position of the target vehicle.

[0012] A method for navigating a host vehicle is provided without being inconsistent with the disclosed embodiments. The method may include the step of receiving a plurality of images representing the host vehicle's environment from an image acquisition device. The method may further include the step of analyzing at least one of the plurality of images to identify a target vehicle in the host vehicle's environment. The method may further include the step of receiving map information associated with the host vehicle's environment. The method may further include the step of determining the trajectory of the target vehicle over a period of time based on the analysis of the plurality of images. The method may further include the step of determining the position of the target vehicle relative to roads in the host vehicle's environment based on the determined trajectory of the target vehicle and the map information. The method may further include the step of determining a navigation action for the host vehicle based on the determined position of the target vehicle.

[0013] Without conflicting with other disclosed embodiments, a non - transient computer - readable storage medium may store program instructions that are executed by at least one processing device and perform any of the methods described herein.

[0014] The above general description and the following detailed description are exemplary and explanatory only and are not restrictive of the claims.

Brief Description of the Drawings

[0015] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate various disclosed embodiments. The drawings are as follows.

[0016] [Figure 1] A schematic diagram of an exemplary system that does not conflict with the disclosed embodiments.

[0017] [Figure 2A] A schematic side view of an exemplary vehicle including a system that does not conflict with the disclosed embodiments.

[0018] [Figure 2B] A schematic plan view of the vehicle and system shown in FIG. 2A that does not conflict with the disclosed embodiments.

[0019] [Figure 2C] A schematic plan view of another embodiment of a vehicle including a system that does not conflict with the disclosed embodiments.

[0020] [Figure 2D] A schematic plan view of yet another embodiment of a vehicle including a system that does not conflict with the disclosed embodiments.

[0021] [Figure 2E] A schematic plan view of yet another embodiment of a vehicle including a system that does not conflict with the disclosed embodiments.

[0022] [Figure 2F] This is a schematic diagram of an exemplary vehicle control system that is consistent with the disclosed embodiments.

[0023] [Figure 3A] This is a schematic diagram of the interior of a vehicle, including a rearview mirror and user interface for a vehicle imaging system, which is consistent with the disclosed embodiments.

[0024] [Figure 3B] This is a diagram of an example of a camera mount configured to be positioned behind the rearview mirror and facing the vehicle's windshield, consistent with the disclosed embodiments.

[0025] [Figure 3C] Figure 3B shows the camera mount from a different perspective that is consistent with the disclosed embodiments.

[0026] [Figure 3D] This is a diagram of an example of a camera mount configured to be positioned behind the rearview mirror and facing the vehicle's windshield, consistent with the disclosed embodiments.

[0027] [Figure 4] This is an exemplary block diagram of memory configured to store instructions for performing one or more operations consistent with the disclosed embodiments.

[0028] [Figure 5A] This flowchart shows an exemplary process for performing one or more navigation responses based on monocular image analysis consistent with the disclosed embodiments.

[0029] [Figure 5B] This flowchart shows an exemplary process for detecting one or more vehicles and / or pedestrians in a set of images consistent with the disclosed embodiments.

[0030] [Figure 5C] This flowchart shows an exemplary process for detecting road markings and / or lane shape information within a set of images consistent with the disclosed embodiments.

[0031] [Figure 5D] This flowchart shows an exemplary process for detecting traffic signal lights in a set of images consistent with the disclosed embodiments.

[0032] [Figure 5E] This flowchart shows an exemplary process for providing one or more navigation responses based on a vehicle path consistent with the disclosed embodiments.

[0033] [Figure 5F] This flowchart shows an exemplary process for determining whether a preceding vehicle is changing lanes, consistent with the disclosed embodiments.

[0034] [Figure 6] This flowchart shows an exemplary process for performing one or more navigation responses based on stereo image analysis consistent with the disclosed embodiments.

[0035] [Figure 7] This flowchart shows an exemplary process for making one or more navigation responses based on the analysis of three sets of images consistent with the disclosed embodiments.

[0036] [Figure 8] A sparse map is shown for providing navigation for an autonomous vehicle that is consistent with the disclosed embodiments.

[0037] [Figure 9A] A polynomial representation of a portion of a road segment, consistent with the disclosed embodiments, is shown.

[0038] [Figure 9B] The diagram shows a curve in three-dimensional space representing the target trajectory of a vehicle for a specific road segment included in a sparse map that is consistent with the disclosed embodiments.

[0039] [Figure 10] Exemplary landmarks that may be included in a sparse map consistent with the disclosed embodiments are shown.

[0040] [Figure 11A] A polynomial representation of the trajectory that is consistent with the disclosed embodiments is shown.

[0041] [Figure 11B] This shows a target trajectory along a multi-lane road that is consistent with the disclosed embodiments. [Figure 11C] This shows a target trajectory along a multi-lane road that is consistent with the disclosed embodiments.

[0042] [Figure 11D] An exemplary road signature profile consistent with the disclosed embodiments is shown.

[0043] [Figure 12] This is a schematic diagram of a system that uses crowdsourced data received from multiple vehicles for autonomous vehicle navigation, consistent with the disclosed embodiments.

[0044] [Figure 13] An exemplary road navigation model for an autonomous vehicle, represented by multiple three-dimensional splines consistent with the disclosed embodiments, is shown.

[0045] [Figure 14] An overview of the map generated by combining location information from a lot of driving, consistent with the disclosed embodiments, is shown.

[0046] [Figure 15]An example of aligning two operations longitudinally is shown, using exemplary signs as landmarks consistent with the disclosed embodiments.

[0047] [Figure 16] An example of aligning many routes longitudinally is shown, using exemplary signs as landmarks that are consistent with the disclosed embodiments.

[0048] [Figure 17] This is a schematic diagram of a system for generating driving data using a camera, vehicle, and server that are consistent with the disclosed embodiments.

[0049] [Figure 18] This is a schematic diagram of a system for crowdsourcing sparse maps that are consistent with the disclosed embodiments.

[0050] [Figure 19] A flowchart illustrating an exemplary process for generating a sparse map for autonomous vehicle navigation along road segments consistent with the disclosed embodiments.

[0051] [Figure 20] A block diagram of the server, consistent with the disclosed embodiments, is shown.

[0052] [Figure 21] A memory block diagram consistent with the disclosed embodiments is shown.

[0053] [Figure 22] This describes a process for clustering vehicle trajectories associated with vehicles that are consistent with the disclosed embodiments.

[0054] [Figure 23] This document describes a vehicle navigation system that may be used for autonomous navigation, consistent with the disclosed embodiments.

[0055] [Figure 24] A memory block diagram consistent with the disclosed embodiments is shown.

[0056] [Figure 25A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0057] [Figure 25B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0058] [Figure 26] A flowchart illustrating exemplary processes for navigating a host vehicle, consistent with the disclosed embodiments, is provided.

[0059] [Figure 27A] Two exemplary splines in three-dimensional coordinates, consistent with the disclosed embodiments, are shown.

[0060] [Figure 27B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0061] [Figure 28A] Two exemplary splines in three-dimensional coordinates, consistent with the disclosed embodiments, are shown.

[0062] [Figure 28B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0063] [Figure 29A] An exemplary spline in two-dimensional coordinates is shown, consistent with the disclosed embodiments.

[0064] [Figure 29B] An exemplary spline in three-dimensional coordinates is shown, consistent with the disclosed embodiments.

[0065] [Figure 30] An exemplary spline projected onto two-dimensional coordinates, consistent with the disclosed embodiments, is shown.

[0066] [Figure 31] The results of the Delaunay triangulation are shown, consistent with the disclosed embodiments.

[0067] [Figure 32] The relevant triangles projected onto the image are shown based on camera positioning, consistent with the disclosed embodiments.

[0068] [Figure 33] A flowchart illustrating exemplary processes for navigating a host vehicle, consistent with the disclosed embodiments, is provided.

[0069] [Figure 34A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0070] [Figure 34B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0071] [Figure 35A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0072] [Figure 35B] Exemplary received images at t1 and t2, consistent with the disclosed embodiments, are shown.

[0073] [Figure 36A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0074] [Figure 36B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0075] [Figure 37] A flowchart illustrating exemplary processes for navigating a host vehicle, consistent with the disclosed embodiments, is provided.

[0076] [Figure 38A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0077] [Figure 38B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0078] [Figure 38C] A diagram is shown illustrating the relationship between the position (x) and width (w) of the target vehicle in the received image, which is consistent with the disclosed embodiments.

[0079] [Figure 39A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0080] [Figure 39B] Exemplary received images are shown that are not inconsistent with the disclosed embodiments.

[0081] [Figure 40A] This shows exemplary positions of the host vehicle relative to road objects, consistent with the disclosed embodiments.

[0082] [Figure 40B] An exemplary received image at t1, consistent with the disclosed embodiments, is shown.

[0083] [Figure 40C] An exemplary received image at t2, consistent with the disclosed embodiments, is shown.

[0084] [Figure 41] A flowchart illustrating exemplary processes for navigating a host vehicle, consistent with the disclosed embodiments, is provided.

[0085] [Figure 42] A diagram illustrating an exemplary driving scenario in which the distance to a detected target vehicle can be determined is shown. [Modes for carrying out the invention]

[0086] The following detailed description refers to the accompanying drawings. Wherever possible, the same reference numerals are used in the drawings and the following description to refer to the same or similar parts. While several exemplary embodiments are described herein, modifications, adaptations, and other implementations are possible. For example, components shown in the drawings may be replaced, added, or modified, and exemplary methods described herein may be modified by replacing, rearranging, removing, or adding steps to the disclosed methods. Thus, the following detailed description is not limited to the disclosed embodiments and examples. Rather, the appropriate scope is defined by the appended claims.

[0087] Overview of autonomous vehicles

[0088] As used in this disclosure, the term “autonomous vehicle” means a vehicle capable of performing at least one navigation change without driver input. “Navigation change” means a change in one or more of the vehicle’s steering, braking, or acceleration. To be autonomous, a vehicle does not need to be fully automatic (e.g., without a driver or complete operation without driver input). Rather, autonomous vehicles include those that can operate under driver control during certain periods and without driver control during other periods. Autonomous vehicles may also include vehicles that only control certain aspects of vehicle navigation, such as steering (e.g., maintaining the vehicle’s course between lane constraints), while leaving other aspects to the driver (e.g., braking). In some cases, an autonomous vehicle may handle some or all aspects of the vehicle’s braking, speed control, and / or steering.

[0089] Human drivers generally rely on visual cues and observations to control their vehicles; therefore, traffic infrastructure is formed such that lane markings, traffic signs, and traffic lights are all designed to provide drivers with visual information. Considering these design characteristics of traffic infrastructure, autonomous vehicles may include cameras and processing units that analyze visual information captured from the vehicle's environment. This visual information may include, for example, components of the traffic infrastructure observable by the driver (e.g., lane markings, traffic signs, traffic lights, etc.) and other obstacles (e.g., other vehicles, pedestrians, litter, etc.). Furthermore, autonomous vehicles may also use stored information, such as information that provides a model of the vehicle's environment when navigating. For example, a vehicle may use GPS data, sensor data (e.g., from accelerometers, speed sensors, suspension sensors, etc.), and / or other map data to provide information related to the environment while the vehicle is traveling, and the vehicle (like other vehicles) may use this information to determine its own position on the model.

[0090] In some embodiments of this disclosure, an autonomous vehicle may use information acquired during navigation (e.g., from cameras, GPS devices, accelerometers, speed sensors, suspension sensors, etc.). In other embodiments, an autonomous vehicle may use information acquired from past navigation by the vehicle (or other vehicles) during navigation. In yet another embodiment, an autonomous vehicle may use a combination of information acquired during navigation and information acquired from past navigation. The following sections provide an overview of a system consistent with the embodiments disclosed, followed by an overview of a forward-facing imaging system and a method consistent with said system. Subsequent sections disclose systems and methods for constructing, using, and updating sparse maps for the navigation of an autonomous vehicle.

[0091] System Overview

[0092] Figure 1 is a block diagram representation of System 100 that is consistent with the exemplary embodiments disclosed. System 100 may include various components depending on the requirements of a particular embodiment. In some embodiments, System 100 may include a processing unit 110, an image acquisition unit 120, a position sensor 130, one or more memory units 140, 150, a map database 160, a user interface 170, and a wireless transceiver 172. The processing unit 110 may include one or more processing devices. In some embodiments, the processing unit 110 may include an application processor 180, an image processor 190, or any other suitable processing device. Similarly, the image acquisition unit 120 may include any number of image acquisition devices and components depending on the requirements of a particular application. In some embodiments, the image acquisition unit 120 may include one or more image acquisition devices (e.g., cameras) such as image acquisition device 122, image acquisition device 124, and image acquisition device 126. System 100 may also include a data interface 128 that enables communication between the processing device 110 and the image acquisition device 120. For example, the data interface 128 may include one or more wired and / or wireless links for transmitting image data acquired by the image acquisition device 120 to the processing unit 110.

[0093] The wireless transceiver 172 may include one or more devices configured to exchange transmissions over an air interface to one or more networks (e.g., cellular, internet, etc.) using radio frequencies, infrared frequencies, magnetic fields, or electric fields. The wireless transceiver 172 may transmit and / or receive data using any known standard (e.g., Wi-Fi®, Bluetooth®, Bluetooth Smart®, 802.15.4, ZigBee®, etc.). Such transmissions may include communication from a host vehicle to one or more remotely located servers. Such transmissions may also include communication (one-way or two-way) between the host vehicle and one or more target vehicles in the host vehicle's environment (e.g., to facilitate the coordination of the host vehicle's navigation, taking into consideration or together with target vehicles in the host vehicle's environment), or broadcast transmissions to unspecified recipients in the vicinity of the transmitting vehicle.

[0094] Both the application processor 180 and the image processor 190 may include various types of processing devices. For example, either or both of the application processor 180 and the image processor 190 may include a microprocessor, a preprocessor (e.g., an image preprocessor), a graphics processing unit (GPU), a central processing unit (CPU), support circuitry, a digital signal processor, an integrated circuit, memory, or any other type of device suitable for application execution and image processing and analysis. In some embodiments, the application processor 180 and / or the image processor 190 may include any type of single or multicore processor, a mobile device microcontroller, a central processing unit, etc. Various processing devices may be used, including processors available from manufacturers such as Intel® and AMD®, or GPUs available from manufacturers such as NVIDIA® and ATI®, and may include various architectures (e.g., x86 processor, ARM®, etc.).

[0095] In some embodiments, the application processor 180 and / or the image processor 190 may include one of the EyeQ series of processor chips available from Mobileye®. Each of these processor designs includes multiple processing units having local memory and instruction sets. Such processors may include video inputs for receiving image data from multiple image sensors and may also include video output capabilities. In one example, EyeQ2 uses 90nm-micron technology operating at 332MHz. The EyeQ2 architecture consists of two floating-point, hyperthreaded 32-bit RISC CPUs (MIPS32 34K cores), five Vision-Computing-Engines (VCEs), three Vector-Microcode Processors (VMPs), a Denali-64-bit mobile DDR controller, a 128-bit internal Sonics interconnect, 16-bit dual video input and 18-bit video output controllers, 16-channel DMA, and several peripherals. The MIPS34K CPU manages five VCEs, three VMPs and DMAs, a second MIPS34K CPU, and a multi-channel DMA and other peripherals. The five VCEs, three VMPs and MIPS34K CPUs can perform the intensive visual computations required by the multifunction bundle application. In another example, an EyeQ3, a third-generation processor that is more than six times more powerful than the EyeQ2, may be used in the disclosed embodiment. In yet another example, an EyeQ4 and / or EyeQ5 may be used in the disclosed embodiment. Of course, any new or future EyeQ processing device may be used with the disclosed embodiment.

[0096] Any of the processing devices disclosed herein may be configured to perform a particular function. Configuring a processing device, for example, the EyeQ processor or other controller or microprocessor described herein, to perform a particular function may include programming computer-executable instructions and making those instructions available to the processing device for execution during the operation of the processing device. In some embodiments, configuring a processing device may include directly programming the processing device with architectural instructions. For example, processing devices, such as field-programmable gate arrays (FPGAs) and application-specific integrated circuits (ASICs), may be configured using, for example, one or more hardware description languages ​​(HDLs).

[0097] In other embodiments, configuring a processing device may include storing executable instructions in a memory accessible to the processing device during operation. For example, the processing device may access the memory during operation to retrieve and execute the stored instructions. In any case, a processing device configured to perform sensing, image analysis, and / or navigation functions disclosed herein represents a dedicated hardware-based system controlling multiple hardware-based components of a host vehicle.

[0098] Figure 1 shows two separate processing devices included in the processing unit 110, but more or fewer processing devices may be used. For example, in some embodiments, a single processing device may be used to perform the tasks of the application processor 180 and the image processor 190. In other embodiments, these tasks may be performed by two or more processing devices. Furthermore, in some embodiments, the system 100 may include one or more processing units 110 that do not include other components, such as the image acquisition unit 120.

[0099] The processing unit 110 may have various types of devices. For example, the processing unit 110 may include various devices such as a controller, an image preprocessor, a central processing unit (CPU), a graphics processing unit (GPU), support circuits, a digital signal processor, an integrated circuit, memory, or any other type of device for image processing and analysis. The image preprocessor may include a video processor that captures, digitizes, and processes images from an image sensor. The CPU may have any number of microcontrollers or microprocessors. The GPU may have any number of microcontrollers or microprocessors. The support circuits may be any number of circuits commonly known in the art, including caches, power supplies, clocks, and input-output circuits. The memory may store software that controls the operation of the system when executed by the processor. The memory may include database and image processing software. The memory may have any number of random access memories, read-only memories, flash memories, disk drives, optical memory devices, tape storage, removable storage, and other types of storage devices. In one example, the memory may be separate from the processing unit 110. In another example, the memory may be integrated into the processing unit 110.

[0100] Each memory unit 140, 150 may contain software instructions that, when executed by a processor (e.g., an application processor 180 and / or an image processor 190), can control various aspects of the operation of the system 100. These memory units may include various database and image processing software, as well as trained systems, such as neural networks or deep neural networks. The memory units may include random access memory (RAM), read-only memory (ROM), flash memory, disk drives, optical memory, tape storage, removable storage, and / or any other type of storage device. In some embodiments, the memory units 140, 150 may be separate from the application processor 180 and / or the image processor 190. In other embodiments, these memory units may be integrated into the application processor 180 and / or the image processor 190.

[0101] The position sensor 130 may include any type of device suitable for determining the position associated with at least one component of the system 100. In some embodiments, the position sensor 130 may include a GPS receiver. Such a receiver can determine the user's position and speed by processing signals broadcast by Global Positioning System satellites. Position information from the position sensor 130 may be available to the application processor 180 and / or the image processor 190.

[0102] In some embodiments, the system 100 may include components such as speed sensors (e.g., tachometer, speedometer) for measuring the speed of the vehicle 200, and / or accelerometers (either single-axis or multi-axis) for measuring the acceleration of the vehicle 200.

[0103] The user interface 170 may include any device suitable for providing information to one or more users of the system 100 or for receiving input from one or more users. In some embodiments, the user interface 170 may include a user input device, such as a touchscreen, microphone, keyboard, pointer device, track wheel, camera, knob, or buttons. Using such an input device, a user may provide information input or commands to the system 100 by typing instructions or information, providing voice commands, using buttons, pointers or eye-tracking functions, or by selecting menu options on the screen through any other suitable technology for communicating information to the system 100.

[0104] The user interface 170 may include one or more processing devices configured to provide and receive information with the user and, for example, to process information for use by the application processor 180. In some embodiments, such processing devices may recognize and track eye movements, receive and interpret voice commands, and execute instructions to recognize and interpret touches and / or gestures made to the touchscreen in response to keyboard input or menu selections, etc. In some embodiments, the user interface 170 may include a display, a speaker, a haptic device, and / or any other device for providing output information to the user.

[0105] The map database 160 may include any type of database for storing map data useful to system 100. In some embodiments, the map database 160 may include data relating to the location in a reference coordinate system of various items, including roads, water features, geographical features, companies, points of interest, restaurants, gas stations, etc. The map database 160 may store not only the locations of such items but also descriptors associated with those items, including, for example, names associated with any of the stored features. In some embodiments, the map database 160 may be physically located together with other components of system 100. Alternatively or further, the map database 160 or a part thereof may be located separately from other components of system 100 (e.g., processing unit 110). In such embodiments, information from the map database 160 may be downloaded via a wired or wireless data connection to a network (e.g., via a cellular network and / or the Internet, etc.). In some cases, the map database 160 may store a sparse data model including a polynomial representation consisting of specific road features (e.g., lane markings) or target trajectories relative to a host vehicle. Systems and methods for generating such maps are discussed below with reference to Figures 8 to 19.

[0106] The image acquisition devices 122, 124, and 126 may each include any type of device suitable for acquiring at least one image from the environment. Furthermore, any number of image acquisition devices may be used to acquire images for input to an image processor. In some embodiments, only a single image acquisition device may be included, while in other embodiments, two, three, or even four or more image acquisition devices may be included. The image acquisition devices 122, 124, and 126 will be further described below with reference to Figures 2B to 2E.

[0107] System 100 or its various components may be incorporated into various different platforms. In some embodiments, System 100 may be included in a vehicle 200, as shown in Figure 2A. For example, the vehicle 200 may comprise either the processing unit 110 or any of the other components of System 100, as described above in relation to Figure 1. In some embodiments, the vehicle 200 may comprise only a single image acquisition device (e.g., a camera), while in other embodiments, multiple image acquisition devices may be used, as discussed in relation to Figures 2B to 2E, for example. As shown in Figure 2A, for example, either of the image acquisition devices 122 and 124 of the vehicle 200 may be part of an ADAS (Advanced Driver-Assistance System) imaging set.

[0108] The image acquisition device included in the vehicle 200 as part of the image acquisition unit 120 may be positioned in any suitable location. In some embodiments, as shown in Figures 2A-2E and 3A-3C, the image acquisition device 122 may be positioned near the rearview mirror. This position may provide a line of sight similar to that of the driver of the vehicle 200, which can be helpful in determining what the driver can and cannot see. The image acquisition device 122 may be positioned anywhere near the rearview mirror, but placing the image acquisition device 122 on the driver side of the mirror may be even more helpful in acquiring images representing the driver's field of view and / or line of sight.

[0109] Other positions of the image acquisition unit 120 relative to the image acquisition device may be used. For example, the image acquisition device 124 may be located on or inside the bumper of the vehicle 200. Such a position may be particularly suitable for an image acquisition device having a wide field of view. The line of sight of an image acquisition device located on the bumper may be different from the line of sight of the driver, and therefore the image acquisition device on the bumper and the driver are not always looking at the same object. The image acquisition devices (e.g., image acquisition devices 122, 124 and 126) may be located in other positions. For example, the image acquisition devices may be located on or inside one or both of the side mirrors of the vehicle 200, on the roof of the vehicle 200, on the hood of the vehicle 200, on the trunk of the vehicle 200, on the side of the vehicle 200, mounted behind or in front of any of the windows of the vehicle 200, mounted on or near the light casings on the front and / or rear of the vehicle 200, etc.

[0110] In addition to the image acquisition device, the vehicle 200 may include various other components of the system 100. For example, a processing unit 110 may be included in the vehicle 200, either integrated with the vehicle's engine control unit (ECU) or separately from the ECU. The vehicle 200 may also include a position sensor 130, such as a GPS receiver, a map database 160, and memory units 140 and 150.

[0111] As previously discussed, the wireless transceiver 172 may transmit and / or receive data over one or more networks (e.g., a cellular network, the Internet, etc.). For example, the wireless transceiver 172 may upload data collected by system 100 to one or more servers and download data from one or more servers. Through the wireless transceiver 172, system 100 may receive periodic or on-demand updates to data stored in, for example, the map database 160, memory 140, and / or memory 150. Similarly, the wireless transceiver 172 may upload any data from system 100 (e.g., images captured by the image acquisition unit 120, data received by the position sensor 130 or other sensors, vehicle control system, etc.) and / or any data processed by the processing unit 110 to one or more servers.

[0112] System 100 may upload data to a server (e.g., the cloud) based on privacy level settings. For example, System 100 may implement privacy level settings to regulate or restrict certain types of data (including metadata) sent to a server that can uniquely identify a vehicle and / or the driver / owner of the vehicle. Such settings may be configured by a user, for example, via a wireless transceiver 172, and may be initialized by factory default settings or by data received by the wireless transceiver 172.

[0113] In some embodiments, system 100 may upload data according to a “high” privacy level, and under the configured settings, system 100 may transmit data (e.g., location information related to routes, captured images, etc.) without using detailed information about a specific vehicle and / or driver / owner. For example, when uploading data according to a “high” privacy setting, system 100 may not include the vehicle identification number (VIN) or the name of the vehicle's driver or owner, and instead may transmit data, e.g., captured images related to routes and / or limited location information.

[0114] Other privacy levels are possible. For example, system 100 may transmit data to the server according to a “medium” privacy level and may include additional information not included under a “high” privacy level, such as the vehicle’s manufacturer and / or model and / or vehicle type (e.g., passenger car, sports utility vehicle, truck, etc.). In some embodiments, system 100 may upload data according to a “low” privacy level. Under a “low” privacy level setting, system 100 may upload data that includes enough information to uniquely identify a particular vehicle, its owner / driver and / or part or all of the route the vehicle is traveling. Such “low” privacy level data may include, for example, one or more of the following: VIN, driver / owner’s name, the vehicle’s starting point before departure, the vehicle’s intended destination, the vehicle’s manufacturer and / or model, and the vehicle type.

[0115] Figure 2A is a schematic side view of an exemplary vehicle imaging system consistent with the disclosed embodiment. Figure 2B is an illustrative schematic top view of the embodiment shown in Figure 2A. As shown in Figure 2B, the disclosed embodiment may include a vehicle 200 having a system 100 in its body, which includes a first image acquisition device 122 located near the rearview mirror and / or near the driver of the vehicle 200, a second image acquisition device 124 located on or within the bumper area of ​​the vehicle 200 (e.g., one of the bumper areas 210), and a processing unit 110.

[0116] As shown in Figure 2C, both image acquisition devices 122 and 124 may be positioned near the rearview mirror and / or near the driver of the vehicle 200. Furthermore, although two image acquisition devices 122 and 124 are shown in Figures 2B and 2C, it should be understood that other embodiments may include more than two image acquisition devices. For example, in the embodiments shown in Figures 2D and 2E, first, second and third image acquisition devices 122, 124 and 126 are included in the system 100 of the vehicle 200.

[0117] As shown in Figure 2D, the image acquisition device 122 may be located near the rearview mirror of the vehicle 200 and / or near the driver, and the image acquisition devices 124 and 126 may be located on or within the bumper area of ​​the vehicle 200 (for example, one of the bumper areas 210). As shown in Figure 2E, the image acquisition devices 122, 124 and 126 may be located near the rearview mirror of the vehicle 200 and / or near the driver's seat. The disclosed embodiments are not limited to a specific number and configuration of image acquisition devices, and the image acquisition devices may be located within and / or at any suitable location on the vehicle 200.

[0118] It should be understood that the disclosed embodiments are not limited to vehicles and may be applicable to other contexts. It should also be understood that the disclosed embodiments are not limited to a specific type of vehicle 200 and may be applicable to any type of vehicle, including automobiles, trucks, trailers and other types of vehicles.

[0119] The first image acquisition device 122 may include any preferred type of image acquisition device. The image acquisition device 122 may include an optical axis. In one example, the image acquisition device 122 may include an Aptina M9V024 WVGA sensor with a global shutter. In other embodiments, the image acquisition device 122 may provide a resolution of 1280 × 960 pixels and may include a rolling shutter. The image acquisition device 122 may include various optical elements. In some embodiments, for example, one or more lenses may be included to provide the image acquisition device with a desired focal length and field of view. In some embodiments, the image acquisition device 122 may be associated with a 6 mm lens or a 12 mm lens. In some embodiments, the image acquisition device 122 may be configured to acquire an image having a desired field of view (FOV) 202, as shown in Figure 2D. For example, the image acquisition device 122 may be configured to have a standard FOV, including, for example, a 46-degree FOV, a 50-degree FOV, a 52-degree FOV, or a larger FOV, within the range of 40 to 56 degrees. Alternatively, the image acquisition device 122 may be configured to have a narrow FOV in the range of 23 to 40 degrees, for example, a 28-degree FOV or a 36-degree FOV. Furthermore, the image acquisition device 122 may be configured to have a wide FOV in the range of 100 to 180 degrees. In some embodiments, the image acquisition device 122 may include a wide-angle bumper camera or a camera having an FOV of up to 180 degrees. In some embodiments, the image acquisition device 122 may be a 7.2M pixel image acquisition device with an aspect ratio of approximately 2:1 (e.g., HxV = 3800 × 1900 pixels) using a horizontal FOV of approximately 100 degrees. Such an image acquisition device may be used instead of a three-image acquisition device configuration. Due to significant lens distortion, the vertical field of view (FOV) of such an image acquisition device can be significantly smaller than 50 degrees in implementations where the image acquisition device uses a radially symmetric lens. For example, such a lens does not need to be radially symmetric to enable a vertical FOV greater than 50 degrees using a horizontal FOV of 100 degrees.

[0120] The first image acquisition device 122 may acquire a plurality of first images related to a scene associated with the vehicle 200. Each of the plurality of first images may be acquired as a series of image scan lines that can be captured using a rolling shutter. Each scan line may contain a plurality of pixels.

[0121] The first image acquisition device 122 may have a scanning speed associated with the acquisition of each of the first series of image scan lines. The scanning speed may refer to the speed at which the image sensor can acquire image data associated with each pixel included in a particular scan line.

[0122] The image acquisition devices 122, 124, and 126 may include any preferred type and number of image sensors, for example, CCD sensors or CMOS sensors. In one embodiment, a CMOS image sensor may be used with a rolling shutter, and each pixel in a row is read one at a time, with the row scanning proceeding row by row until the entire image frame is acquired. In some embodiments, rows may be acquired sequentially from top to bottom of the frame.

[0123] In some embodiments, one or more of the image acquisition devices disclosed herein (e.g., image acquisition devices 122, 124, and 126) may constitute a high-resolution camera and may have a resolution greater than 5 megapixels, 7 megapixels, 10 megapixels, or larger pixels.

[0124] The use of a rolling shutter can result in pixels in different rows being exposed and captured at different times, potentially introducing skew and other image artifacts into the captured image frame. In contrast, with an image acquisition device 122 configured to operate with a global or synchronous shutter, all pixels may be exposed for the same amount of time during a common exposure period. Consequently, image data within a frame collected from a system using a global shutter represents a snapshot of the entire FOV at a specific time (e.g., FOV 202). In contrast, when a rolling shutter is applied, each row in the frame is exposed, and data is captured at different times. Therefore, distortion may appear in moving objects in an image acquisition device with a rolling shutter. This phenomenon is described in more detail below.

[0125] The second image acquisition device 124 and the third image acquisition device 126 may be any type of image acquisition device. Like the first image acquisition device 122, each of the image acquisition devices 124 and 126 may include an optical axis. In one embodiment, each of the image acquisition devices 124 and 126 may include an Aptina M9V024 WVGA sensor with a global shutter. Alternatively, each of the image acquisition devices 124 and 126 may include a rolling shutter. Like the image acquisition device 122, the image acquisition devices 124 and 126 may be configured to include various lenses and optical elements. In some embodiments, the lenses associated with the image acquisition devices 124 and 126 may provide the same or narrower FOV (e.g., FOV 204 and 206) as (e.g., FOV 202 associated with image acquisition device 122). For example, the image acquisition devices 124 and 126 may have an FOV of 40 degrees, 30 degrees, 26 degrees, 23 degrees, 20 degrees, or lower.

[0126] Image acquisition devices 124 and 126 may acquire a plurality of second and third images associated with a scene associated with the vehicle 200. Each of the plurality of second and third images may be acquired as a second and third series of image scan lines, which can be captured using a rolling shutter. Each scan line or line may have a plurality of pixels. Image acquisition devices 124 and 126 may have second and third scanning speeds associated with the acquisition of each of the image scan lines included in the second and third series.

[0127] Each image acquisition device 122, 124, and 126 can be positioned in any suitable location and orientation relative to the vehicle 200. The relative arrangement of the image acquisition devices 122, 124, and 126 may be selected to facilitate the combination of information acquired from the image acquisition devices. For example, in some embodiments, the FOV associated with image acquisition device 124 (e.g., FOV 204) may partially or completely overlap with the FOV associated with image acquisition device 122 (e.g., FOV 202) and the FOV associated with image acquisition device 126 (e.g., FOV 206).

[0128] The image acquisition devices 122, 124, and 126 may be positioned in the vehicle 200 at any appropriate relative height. In one example, there may be height differences between the image acquisition devices 122, 124, and 126, which can provide sufficient parallax information to enable stereo analysis. For example, as shown in Figure 2A, the two image acquisition devices 122 and 124 are at different heights. There may also be lateral displacement differences between the image acquisition devices 122, 124, and 126, which, for example, provide additional parallax information for stereo analysis by the processing unit 110. As shown in Figures 2C and 2D, the difference in lateral displacement is d xThis can be shown by [this]. In some embodiments, a forward or backward offset (e.g., range offset) may exist between the image acquisition devices 122, 124, and 126. For example, image acquisition device 122 may be positioned 0.5 to 2 meters or more behind image acquisition devices 124 and / or image acquisition device 126. This type of offset may allow one of the image acquisition devices to cover a possible blind spot of the other image acquisition devices.

[0129] The image acquisition device 122 may have any suitable resolution function (e.g., the number of pixels associated with the image sensor), and the resolution of the image sensor associated with the image acquisition device 122 may be higher, lower, or the same as the resolution of the image sensors associated with the image acquisition devices 124 and 126. In some embodiments, the image sensors associated with the image acquisition device 122 and / or the image acquisition devices 124 and 126 may have a resolution of 640×480, 1024×768, 1280×960, or any other suitable resolution.

[0130] The frame rate (for example, the rate at which an image acquisition device acquires a set of pixel data for one image frame before moving on to acquiring the pixel data associated with the next image frame) may be controllable. The frame rate associated with image acquisition device 122 may be higher, lower, or the same as the frame rates associated with image acquisition devices 124 and 126. The frame rates associated with image acquisition devices 122, 124, and 126 may depend on various factors that may affect the timing of the frame rate. For example, one or more of the image acquisition devices 122, 124, and 126 may include a selectable pixel delay time before and after acquiring image data associated with one or more pixels of the image sensor in image acquisition devices 122, 124, and / or 126. In general, the image data corresponding to each pixel may be acquired according to the clock rate for the device (for example, a clock cycle per pixel). Furthermore, in embodiments including a rolling shutter, one or more of the image acquisition devices 122, 124, and 126 may include a selectable horizontal blanking period before and after acquiring image data associated with rows of pixels of the image sensor in the image acquisition devices 122, 124, and / or 126. Furthermore, one or more of the image acquisition devices 122, 124, and / or 126 may include a selectable vertical blanking period before and after acquiring image data associated with image frames of the image acquisition devices 122, 124, and 126.

[0131] These timing controls can enable synchronization of the frame rates associated with image acquisition devices 122, 124, and 126, even if their respective line scanning speeds are different. Furthermore, as will be discussed in more detail below, these selectable timing controls can enable synchronization of image acquisition from regions where the FOV of image acquisition device 122 overlaps with one or more FOVs of image acquisition devices 124 and 126, even if the field of view of image acquisition device 122 differs from the FOVs of image acquisition devices 124 and 126, among other factors (e.g., image sensor resolution, maximum line scanning speed, etc.).

[0132] The frame rate timing in image acquisition devices 122, 124, and 126 may depend on the resolution of the associated image sensors. For example, assuming similar line scanning speeds for both devices, if one device includes an image sensor with a resolution of 640 × 480 and the other device includes an image sensor with a resolution of 1280 × 960, more time will be required to acquire frames of image data from the sensor with the higher resolution.

[0133] Another factor that may affect the timing of image data acquisition in image acquisition devices 122, 124, and 126 is the maximum line scan speed. For example, acquiring a row of image data from the image sensors included in image acquisition devices 122, 124, and 126 requires a certain minimum amount of time. Assuming no pixel delay time is added, this minimum amount of time to acquire one row of image data is related to the maximum line scan speed for a particular device. Devices offering higher maximum line scan speeds may provide a higher frame rate than devices with lower maximum line scan speeds. In some embodiments, one or more of the image acquisition devices 124 and 126 may have a maximum line scan speed higher than the maximum line scan speed associated with image acquisition device 122. In some embodiments, the maximum line scan speed of image acquisition devices 124 and / or 126 may be 1.25 times, 1.5 times, 1.75 times, 2 times, or greater than the maximum line scan speed of image acquisition device 122.

[0134] In another embodiment, the image acquisition devices 122, 124, and 126 may have the same maximum line scanning speed, but image acquisition device 122 may operate at a scanning speed less than or equal to its maximum scanning speed. The system may be configured so that one or more of the image acquisition devices 124 and 126 operate at a line scanning speed equal to the line scanning speed of image acquisition device 122. In other cases, the system may be configured so that the line scanning speeds of image acquisition device 124 and / or image acquisition device 126 can be 1.25 times, 1.5 times, 1.75 times, 2 times, or more than the line scanning speed of image acquisition device 122.

[0135] In some embodiments, the image acquisition devices 122, 124, and 126 may be asymmetrical; that is, they may include cameras with varying fields of view (FOV) and focal lengths. For example, the fields of view of the image acquisition devices 122, 124, and 126 may include any desired area of ​​the environment of the vehicle 200. In some embodiments, one or more of the image acquisition devices 122, 124, and 126 may be configured to acquire image data from the environment in front of the vehicle 200, behind the vehicle 200, to the sides of the vehicle 200, or a combination thereof.

[0136] Furthermore, the focal lengths associated with each image acquisition device 122, 124, and / or 126 may be selectable (for example, by including appropriate lenses, etc.) so that each device acquires images of objects within a desired distance range relative to the vehicle 200. For example, in some embodiments, the image acquisition devices 122, 124, and 126 may acquire images of objects approaching within a few meters of the vehicle. The image acquisition devices 122, 124, and 126 may be configured to acquire images of objects at a greater distance from the vehicle (for example, 25m, 50m, 100m, 150m, or further). Furthermore, the focal lengths of the image acquisition devices 122, 124, and 126 can be selected so that one image acquisition device (e.g., image acquisition device 122) can acquire images of objects relatively close to the vehicle (e.g., within 10m or 20m), while the other image acquisition devices (e.g., image acquisition devices 124 and 126) can acquire images of objects further away from the vehicle 200 (e.g., further than 20m, 50m, 100m, 150m, etc.).

[0137] According to some embodiments, the field of view (FOV) of one or more image acquisition devices 122, 124, and 126 may be wide-angle. For example, a 140-degree FOV is preferable for image acquisition devices 122, 124, and 126 that may be used to acquire images of areas near the vehicle 200. For example, image acquisition device 122 may be used to acquire images of areas to the right or left of the vehicle 200, and in such embodiments, it may be desirable for image acquisition device 122 to have a wide FOV (e.g., at least 140 degrees).

[0138] The fields of view associated with each of the image acquisition devices 122, 124, and 126 may depend on their respective focal lengths. For example, as the focal length increases, the corresponding field of view shrinks.

[0139] Image acquisition devices 122, 124, and 126 can be configured to have any suitable field of view. In one particular example, image acquisition device 122 may have a horizontal FOV of 46 degrees, image acquisition device 124 may have a horizontal FOV of 23 degrees, and image acquisition device 126 may have a horizontal FOV between 23 degrees and 46 degrees. In another example, image acquisition device 122 may have a horizontal FOV of 52 degrees, image acquisition device 124 may have a horizontal FOV of 26 degrees, and image acquisition device 126 may have a horizontal FOV between 26 degrees and 52 degrees. In some embodiments, the ratio of the FOV of image acquisition device 122 to the FOV of image acquisition device 124 and / or image acquisition device 126 may vary from 1.5 to 2.0. In other embodiments, this ratio may vary between 1.25 and 2.25.

[0140] System 100 may be configured such that the field of view of image acquisition device 122 overlaps at least partially or completely with the field of view of image acquisition device 124 and / or image acquisition device 126. In some embodiments, System 100 may be configured such that the fields of view of image acquisition devices 124 and 126 are, for example, within the range of the field of view of image acquisition device 122 (e.g., narrower) and share a common center with the field of view of image acquisition device 122. In other embodiments, image acquisition devices 122, 124 and 126 may capture adjacent FOVs or have portions that partially overlap in their FOVs. In some embodiments, the fields of view of image acquisition devices 122, 124 and 126 may be aligned such that the centers of the narrow-FOV image acquisition device 124 and / or 126 are located in the lower half of the field of view of the wide-FOV device 122.

[0141] Figure 2F is a schematic diagram of an exemplary vehicle control system consistent with the disclosed embodiments. As shown in Figure 2F, the vehicle 200 may include a throttling system 220, a braking system 230, and a steering system 240. System 100 may provide inputs (e.g., control signals) to one or more of the throttling system 220, brake system 230, and steering system 240 via one or more data links (e.g., any one or more wired and / or wireless links for transmitting data). For example, based on the analysis of images acquired by image acquisition devices 122, 124, and / or 126, System 100 may provide control signals to one or more of the throttling system 220, brake system 230, and steering system 240 to navigate the vehicle 200 (e.g., by accelerating, turning, changing lanes, etc.). Furthermore, System 100 may receive inputs from one or more of the throttling system 220, brake system 230, and steering system 240 indicating the operating conditions of the vehicle 200 (e.g., speed, whether the vehicle 200 is braking and / or turning, etc.). Further details are provided below in reference to Figures 4 to 7.

[0142] As shown in Figure 3A, the vehicle 200 may include a user interface 170 for interacting with the driver or passenger of the vehicle 200. For example, the user interface 170 applied to the vehicle may include a touchscreen 320, a knob 330, buttons 340, and a microphone 350. The driver or passenger of the vehicle 200 may interact with the system 100 using a handle (for example, located on or near the steering column of the vehicle 200, including a turn signal handle) and buttons (for example, located on the steering wheel of the vehicle 200). In some embodiments, the microphone 350 may be located adjacent to the rearview mirror 310. Similarly, in some embodiments, the image acquisition device 122 may be located near the rearview mirror 310. In some embodiments, the user interface 170 may include one or more speakers 360 (for example, speakers of the vehicle's audio system). For example, the system 100 may provide various notifications (for example, alerts) via the speakers 360.

[0143] Figures 3B to 3D illustrate an exemplary camera mount 370 configured to be positioned behind a rearview mirror (e.g., rearview mirror 310) and facing the windshield of a vehicle, consistent with the disclosed embodiments. As shown in Figure 3B, the camera mount 370 may include image acquisition devices 122, 124, and 126. The image acquisition devices 124 and 126 may be positioned behind a glare shield 380, which may fit snugly onto the windshield of a vehicle and include a film and / or anti-reflective material configuration. For example, the glare shield 380 may be positioned to align the shield with the windshield of a vehicle having a matching inclination. In some embodiments, each of the image acquisition devices 122, 124, and 126 may be positioned behind the glare shield 380, as shown, for example, in Figure 3D. The disclosed embodiments are not limited to the specific configurations of the image acquisition devices 122, 124, and 126, the camera mount 370, and the glare shield 380. Figure 3C is a front view of the camera mount 370 shown in Figure 3B.

[0144] As will be understood by those skilled in the art who have access to this disclosure, numerous modifications and / or alterations may be made to the above-disclosed embodiments. For example, not all components are necessarily essential for the operation of system 100. Furthermore, any component may be placed in any suitable part of system 100, and such component may be reconfigured into various configurations while providing the functionality of the disclosed embodiments. Thus, the above configurations are examples, and notwithstanding the above configurations, system 100 can provide a wide range of functionality, such as analyzing the periphery of vehicle 200 and navigating vehicle 200 in response to the analysis.

[0145] System 100 may provide various features related to autonomous driving and / or driver assistance technologies, as will be discussed in more detail below and in accordance with the various disclosed embodiments. For example, System 100 may analyze image data, location data (e.g., GPS location information), map data, speed data, and / or data from sensors included in the vehicle 200. System 100 may collect data for analysis from, for example, an image acquisition unit 120, a location sensor 130, and other sensors. Furthermore, System 100 may analyze the collected data to determine whether the vehicle 200 should perform a particular action, and then automatically perform the determined action without human intervention. For example, if the vehicle 200 is to navigate without human intervention, System 100 may automatically control the braking, acceleration, and / or steering of the vehicle 200 (for example, by transmitting control signals to one or more of the throttling system 220, brake system 230, and steering system 240). Furthermore, the system 100 may analyze the collected data and, based on the analysis of the collected data, issue warnings and / or alerts to the vehicle's occupants. Further details regarding the various embodiments provided by the system 100 are provided below.

[0146] Forward-looking multi-imaging system

[0147] As described above, system 100 may provide a driver assistance function using a multi-camera system. The multi-camera system may use one or more cameras facing the direction of travel of the vehicle. In other embodiments, the multi-camera system may include one or more cameras facing the side or rear of the vehicle. In one embodiment, for example, system 100 may use two camera imaging systems, where a first camera and a second camera (e.g., image acquisition devices 122 and 124) may be positioned in front of and / or to the side of the vehicle (e.g., vehicle 200). The first camera may have a field of view wider, narrower, or partially overlapping with that of the second camera. Furthermore, the first camera may be connected to a first image processor that performs monocular image analysis of the images provided by the first camera, and the second camera may be connected to a second image processor that performs monocular image analysis of the images provided by the second camera. The outputs of the first and second image processors (e.g., processed information) may be combined. In some embodiments, a second image processor may receive images from both the first and second cameras and perform stereo analysis. In another embodiment, system 100 may use a three-camera imaging system in which each camera has a different field of view. Thus, such a system may make decisions based on information derived from objects located at various distances both in front of and to the side of the vehicle. Reference to monocular image analysis may refer to an example in which image analysis is performed based on images acquired from a single viewpoint (e.g., from a single camera). Stereo image analysis may refer to an example in which image analysis is performed based on two or more images acquired with one or more variations of image acquisition parameters. For example, acquired images suitable for performing stereo image analysis may include images acquired from two or more different positions, from various fields of view, using different focal lengths, along with disparity information.

[0148] For example, in one embodiment, system 100 may implement a three-camera configuration using image acquisition devices 122, 124, and 126. In such a configuration, image acquisition device 122 may provide a narrow field of view (e.g., 34 degrees, or other values ​​selected from a range such as about 20 to 45 degrees), image acquisition device 124 may provide a wide field of view (e.g., 150 degrees, or other values ​​selected from a range such as about 100 to about 180 degrees), and image acquisition device 126 may provide an intermediate field of view (e.g., 46 degrees, or other values ​​selected from a range such as about 35 to about 60 degrees). In some embodiments, image acquisition device 126 may operate as a main camera or primary camera. Image acquisition devices 122, 124, and 126 may be positioned behind the rearview mirror 310 and substantially side by side (e.g., 6 cm apart). Furthermore, in some embodiments, as described above, one or more of the image acquisition devices 122, 124, and 126 may be mounted behind a glare shield 380 that closely overlaps the windshield of the vehicle 200. Such a shield may operate to minimize the influence of any reflections from inside the vehicle on the image acquisition devices 122, 124, and 126.

[0149] In another embodiment, as described above in relation to Figures 3B and 3C, a wide-field camera (e.g., image acquisition device 124 in the above example) may be mounted below the narrow and primary-field cameras (e.g., image devices 122 and 126 in the above example). This configuration may provide a free line of sight from the wide-field camera. To reduce reflections, the camera may be mounted near the windshield of the vehicle 200, and the camera may include polarizers to attenuate reflected light.

[0150] A three-camera system can offer specific performance characteristics. For example, some embodiments may include the ability to verify object detection by one camera based on detection results from another camera. In the three-camera configuration described above, the processing unit 110 may include three processing devices (e.g., three EyeQ series processor chips as described above), each processing device specializing in processing images captured by one or more of the image acquisition devices 122, 124, and 126.

[0151] In a three-camera system, the first processing device may receive images from both the main camera and the narrow-field-of-view camera, and perform visual processing on the narrow-field-of-view camera to detect, for example, other vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. Furthermore, the first processing device may calculate the pixel parallax between the images from the main camera and the narrow-field-of-view camera and generate a 3D reconstruction of the vehicle 200's environment. The first processing device may then combine the 3D reconstruction with 3D map data or 3D information calculated based on information from another camera.

[0152] A second processing device may receive images from the main camera and perform visual processing to detect other vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. Furthermore, the second processing device may calculate camera displacement and, based on that displacement, calculate pixel parallax between a series of images to generate a 3D reconstruction of the scene (e.g., structure from motion). The second processing device may transmit the structure from motion, based on the 3D reconstruction of the object combined with stereo 3D images, to the first processing device.

[0153] A third processing device may receive images from a wide-field-of-view camera and process the images to detect vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. The third processing device may further analyze the images by executing additional processing commands to identify objects moving within the images, such as vehicles changing lanes or pedestrians.

[0154] In some embodiments, the independent capture and processing of image-based information streams may provide an opportunity to offer redundancy to the system. Such redundancy may include, for example, verifying and / or supplementing information obtained by capturing and processing image information from at least a second image capture device using a first image capture device and images processed from that device.

[0155] In some embodiments, the system 100 may use two image acquisition devices (e.g., image acquisition devices 122 and 124) when providing navigation assistance to the vehicle 200, and a third image acquisition device (e.g., image acquisition device 126) may be used to provide redundancy and verify the analysis of data received from the other two image acquisition devices. For example, in such a configuration, image acquisition devices 122 and 124 may provide images for stereo analysis by the system 100 to navigate the vehicle 200, while image acquisition device 126 may provide images for monocular analysis by the system 100 to provide redundancy and verification of information acquired based on images acquired from image acquisition devices 122 and / or image acquisition devices 124. In other words, image acquisition device 126 (and the corresponding processing device) can be considered as providing a redundant subsystem to provide checks on the analysis derived from image acquisition devices 122 and 124 (for example, to provide an automatic emergency braking (AEB) system). Furthermore, in some embodiments, the redundancy and verification of received data may be supplemented based on information received from one or more sensors (e.g., radar, lidar, acoustic sensors, information received from one or more transceivers outside the vehicle).

[0156] Those skilled in the art will recognize that the above-described camera configurations, camera arrangements, number of cameras, and camera positions are merely examples. These components and other aspects described for the overall system may be assembled and used in a variety of different configurations without departing from the scope of the disclosed embodiments. Further details regarding the use of multi-camera systems to provide driver assistance and / or autonomous vehicle functionality are as follows:

[0157] Figure 4 is an exemplary functional block diagram of memory 140 and / or 150, which may store / program instructions for performing one or more operations consistent with the disclosed embodiments. Hereafter, we will refer to memory 140, but those skilled in the art will recognize that instructions may be stored in memory 140 and / or 150.

[0158] As shown in Figure 4, memory 140 may store a monocular image analysis module 402, a stereo image analysis module 404, a velocity and acceleration module 406, and a navigation response module 408. The disclosed embodiments are not limited to a specific configuration of memory 140. Furthermore, the application processor 180 and / or the image processor 190 may execute instructions stored in any of the modules 402, 404, 406, and 408 contained in memory 140. Those skilled in the art will understand that the following references to the processing unit 110 may refer to the application processor 180 and the image processor 190 individually or collectively. Accordingly, any of the following stages of processing may be performed by one or more processing devices.

[0159] In one embodiment, the monocular image analysis module 402 may store instructions (e.g., computer vision software) to perform monocular image analysis of a set of images acquired by one of the image acquisition devices 122, 124, and 126, when executed by the processing unit 110. In some embodiments, the processing unit 110 may perform monocular image analysis by combining information from the set of images with additional perceptual information (e.g., information from radar, lidar, etc.). As described below with reference to Figures 5A to 5D, the monocular image analysis module 402 may include instructions for detecting a set of features in the image set, such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, hazardous objects, and any other features associated with the vehicle's environment. Based on the analysis, the system 100 (e.g., via the processing unit 110) may cause one or more navigation responses in the vehicle 200, such as a change of direction, lane change, and acceleration change, as discussed below with reference to the navigation response module 408.

[0160] In one embodiment, the stereo image analysis module 404 may store instructions (e.g., computer vision software) to perform stereo image analysis on first and second image sets acquired by a combination of image acquisition devices selected from among image acquisition devices 122, 124, and 126, when executed by the processing unit 110. In some embodiments, the processing unit 110 may perform stereo image analysis by combining information from the first and second image sets with additional perceptual information (e.g., information from radar). For example, the stereo image analysis module 404 may include instructions for performing stereo image analysis based on a first set of images acquired by image acquisition device 124 and a second set of images acquired by image acquisition device 126. As will be described below with reference to Figure 6, the stereo image analysis module 404 may include instructions for detecting a set of features in the first and second image sets, such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, and hazardous objects. Based on the analysis, the processing unit 110 may cause the vehicle 200 to perform one or more navigation responses, such as changes in direction, lane changes, and changes in acceleration, as discussed below in relation to the navigation response module 408. Furthermore, in some embodiments, the stereo image analysis module 404 may implement a technique associated with a system that can be configured to detect and / or classify objects in an environment from which perceptual information has been taken up and processed, using either a trained system (e.g., a neural network or a deep neural network) or an untrained system, such as a computer vision algorithm. In one embodiment, the stereo image analysis module 404 and / or other image processing modules may be configured to use a combination of trained and untrained systems.

[0161] In one embodiment, the velocity and acceleration module 406 may store software configured to analyze data received from one or more computing and electromechanical devices within the vehicle 200 that are configured to change the velocity and / or acceleration of the vehicle 200. For example, the processing unit 110 may execute instructions associated with the velocity and acceleration module 406 to calculate a target velocity for the vehicle 200 based on data derived from the execution of the monocular image analysis module 402 and / or the stereo image analysis module 404. Such data may include, for example, a target position, velocity and / or acceleration, the position and / or velocity of the vehicle 200 relative to nearby vehicles, pedestrians or road objects, and position information of the vehicle 200 relative to road lane markings. Furthermore, the processing unit 110 may calculate a target velocity for the vehicle 200 based on perceptual input (e.g., information from radar) and input from other systems of the vehicle 200, such as the vehicle 200's throttling system 220, brake system 230 and / or steering system 240. Based on the calculated target speed, the processing unit 110 may transmit electronic signals to the vehicle 200's throttling system 220, brake system 230, and / or steering system 240 to trigger changes in speed and / or acceleration, for example, by physically applying the brakes or releasing the accelerator of the vehicle 200.

[0162] In one embodiment, the navigation response module 408 may store software executable by the processing unit 110 to determine a desired navigation response based on data derived from the execution of the monocular image analysis module 402 and / or the stereo image analysis module 404. Such data may include position and velocity information associated with nearby vehicles, pedestrians and road objects, and target position information of the vehicle 200. Furthermore, in some embodiments, the navigation response may be based (partially or entirely) on the relative velocity or relative acceleration between the vehicle 200 and one or more objects detected from the execution of the monocular image analysis module 402 and / or the stereo image analysis module 404. The navigation response module 408 may also determine a desired navigation response based on perceptual input (e.g., information from radar) and input from other systems of the vehicle 200, such as the vehicle 200's throttling system 220, brake system 230 and steering system 240. Based on the desired navigation response, the processing unit 110 may transmit electronic signals to the vehicle 200's throttling system 220, brake system 230, and steering system 240 to trigger the desired navigation response, for example, by rotating the vehicle 200's steering wheel to achieve a predetermined angle of rotation. In some embodiments, the processing unit 110 may use the output of the navigation response module 408 (e.g., the desired navigation response) as input to the operation of the speed and acceleration module 406 in order to calculate the change in the vehicle 200's speed.

[0163] Furthermore, any of the modules disclosed herein (e.g., modules 402, 404, and 406) may implement techniques associated with trained systems (such as neural networks or deep neural networks) or untrained systems.

[0164] Figure 5A is a flowchart illustrating an exemplary process 500A for performing one or more navigation responses based on monocular image analysis, consistent with the disclosed embodiments. In step 510, the processing unit 110 may receive multiple images via a data interface 128 between the processing unit 110 and the image acquisition unit 120. For example, a camera included in the image acquisition unit 120 (e.g., an image acquisition device 122 having a field of view 202) may acquire multiple images of the area in front of the vehicle 200 (or, for example, to the side or rear of the vehicle) and transmit them to the processing unit 110 via a data connection (e.g., digital, wired, USB, wireless, Bluetooth®, etc.). In step 520, the processing unit 110 may perform analysis of the multiple images by running a monocular image analysis module 402, as will be described in more detail below with reference to Figures 5B to 5D. By performing the analysis, the processing unit 110 may detect feature sets within the image set, such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, and traffic signals.

[0165] In step 520, the processing unit 110 may run the monocular image analysis module 402 to detect various road obstacles, such as truck tire parts, fallen road signs, loose cargo, and small animals. Road obstacles vary in structure, shape, size, and color, which can make their detection more difficult. In some embodiments, the processing unit 110 may run the monocular image analysis module 402 and perform multi-frame analysis on multiple images to detect road obstacles. For example, the processing unit 110 may estimate the camera movement between consecutive image frames and calculate the pixel parallax between frames to construct a 3D map of the road. The processing unit 110 may then use the 3D map to detect the road surface and any obstacles present on the road surface.

[0166] In step 530, the processing unit 110 may execute the navigation response module 408 to perform one or more navigation responses in the vehicle 200 based on the analysis performed in step 520 and the techniques described above in relation to Figure 4. Navigation responses may include, for example, a change of direction, a lane change, and a change in acceleration. In some embodiments, the processing unit 110 may perform one or more navigation responses using data derived from the execution of the speed and acceleration module 406. Furthermore, multiple navigation responses may occur simultaneously, sequentially, or in any combination thereof. For example, the processing unit 110 may change the vehicle 200 to the other lane and then accelerate by, for example, transmitting control signals in succession to the steering system 240 and throttling system 220 of the vehicle 200. Alternatively, the processing unit 110 may change lanes while braking the vehicle 200 by, for example, simultaneously transmitting control signals to the brake system 230 and steering system 240 of the vehicle 200.

[0167] Figure 5B is a flowchart illustrating an exemplary process 500B for detecting one or more vehicles and / or pedestrians in a set of images, consistent with the disclosed embodiments. Processing unit 110 may perform process 500B by running monocular image analysis module 402. In step 540, processing unit 110 may determine a set of candidate objects representing possible vehicles and / or pedestrians. For example, processing unit 110 may scan one or more images and compare the images to one or more predetermined patterns to identify locations in each image that may contain objects of interest (e.g., vehicles, pedestrians, or parts thereof). The predetermined patterns may be designed in such a way that a high percentage of "false hits" and a low percentage of "misses" occur. For example, processing unit 110 may use low-threshold similarity with the predetermined patterns to identify candidate objects as possible vehicles or pedestrians. Doing so may reduce the probability that processing unit 110 misses (e.g., fails to identify) candidate objects representing vehicles or pedestrians.

[0168] In step 542, the processing unit 110 may filter the set of candidate objects and exclude certain candidates (e.g., irrelevant or unrelated objects) based on classification criteria. Such criteria may be derived from various characteristics associated with the types of objects stored in a database (e.g., a database stored in memory 140). Characteristics may include the shape, dimensions, texture, and location of an object (e.g., relative to the vehicle 200). Thus, the processing unit 110 may use one or more sets of criteria to reject false candidates from the set of candidate objects.

[0169] In step 544, the processing unit 110 may analyze multiple frames of the image to determine whether an object in the set of candidate objects represents a vehicle and / or a pedestrian. For example, the processing unit 110 may track the detected candidate objects across consecutive frames and accumulate frame-by-frame data associated with the detected object (e.g., size, position relative to the vehicle 200). Furthermore, the processing unit 110 may estimate parameters for the detected object and compare the frame-by-frame position data of the object with the predicted position.

[0170] In step 546, the processing unit 110 may construct a set of measurements for the detected object. Such measurements may include, for example, position, velocity, and acceleration values ​​associated with the detected object (relative to the vehicle 200). In some embodiments, the processing unit 110 may construct measurements based on a series of time-based observations, e.g., estimation techniques using a Kalman filter or linear quadratic estimation (LQE), and / or based on available modeling data for different object types (e.g., cars, trucks, pedestrians, bicycles, road signs, etc.). The Kalman filter may be based on a measurement of the object's scale, where the scale measurement is proportional to the time to collision (e.g., the length of time it takes for the vehicle 200 to reach the object). Thus, by performing steps 540-546, the processing unit 110 may identify vehicles and pedestrians appearing in the captured set of images and derive information associated with the vehicles and pedestrians (e.g., position, velocity, size). Based on the identified and derived information, the processing unit 110 may perform one or more navigation responses in the vehicle 200, as described in relation to Figure 5A above.

[0171] In step 548, the processing unit 110 may perform optical flow analysis on one or more images to reduce the probability of detecting a "false hit" and the probability of missing a candidate object representing a vehicle or pedestrian. Optical flow analysis may refer to, for example, analyzing motion patterns distinct from road surface motion for a vehicle 200 in one or more images associated with other vehicles and pedestrians. The processing unit 110 may calculate the motion of a candidate object by observing the different positions of the object across multiple image frames captured at different times. The processing unit 110 may use position and time values ​​as input to a mathematical model for calculating the motion of the candidate object. Thus, optical flow analysis may provide another method for detecting vehicles and pedestrians near the vehicle 200. The processing unit 110 may perform optical flow analysis in combination with steps 540-546 to provide redundancy for detecting vehicles and pedestrians and improve the reliability of the system 100.

[0172] Figure 5C is a flowchart illustrating an exemplary process 500C for detecting road markings and / or lane shape information in a set of images, consistent with the disclosed embodiments. Processing unit 110 may perform process 500C by running monocular image analysis module 402. In step 550, processing unit 110 may detect a set of objects by scanning one or more images. To detect lane markings, lane shape information and other relevant road marking segments, processing unit 110 may filter the set of objects to exclude those determined to be irrelevant (e.g., small depressions, small rocks). In step 552, processing unit 110 may group together segments detected in step 550 that belong to the same road marking or lane marking. Based on the grouping, processing unit 110 may develop a model, such as a mathematical model, representing the detected segments.

[0173] In step 554, the processing unit 110 may construct a set of measurements associated with the detected segment. In some embodiments, the processing unit 110 may create a projection of the detected segment from the image plane onto a real-world plane. The projection may be characterized using a cubic polynomial with coefficients corresponding to physical properties, such as the detected road position, slope, curvature, and curvature function. When generating the projection, the processing unit 110 may consider changes in the road surface and the pitch and roll rates associated with the vehicle 200. Furthermore, the processing unit 110 may model the elevation of the road by analyzing the position and motion cues present on the road surface. Furthermore, the processing unit 110 may estimate the pitch and roll rates associated with the vehicle 200 by tracking a set of feature points in one or more images.

[0174] In step 556, the processing unit 110 may perform multi-frame analysis, for example, by tracking detected segments across consecutive image frames and accumulating frame-by-frame data associated with the detected segments. As the processing unit 110 performs multi-frame analysis, the set of measurements constructed in step 554 becomes more reliable and can be associated with an increasingly higher level of confidence. Thus, by performing steps 550, 552, 554, and 556, the processing unit 110 can identify road markings appearing in the acquired set of images and derive lane shape information. Based on the identified and derived information, the processing unit 110 may perform one or more navigation responses in the vehicle 200, as described in relation to Figure 5A above.

[0175] In step 558, the processing unit 110 may consider additional information sources to further develop the safety model for the vehicle 200 in its surrounding context. The processing unit 110 may use the safety model to define a context in which the system 100 can perform autonomous control of the vehicle 200 in a safe manner. To develop the safety model, in some embodiments, the processing unit 110 may consider the position and movement of other vehicles, detected road edges and boundaries, and / or general road geometry descriptions extracted from map data (e.g., data from the map database 160). By considering additional information sources, the processing unit 110 may provide redundancy for detecting road markings and lane geometry, thereby increasing the reliability of the system 100.

[0176] Figure 5D is a flowchart illustrating an exemplary process 500D for detecting traffic lights in a set of images, consistent with the disclosed embodiments. Processing unit 110 may perform process 500D by running monocular image analysis module 402. In step 560, processing unit 110 may scan the set of images and identify objects appearing at locations in the images that are likely to contain traffic lights. For example, processing unit 110 may filter the identified objects to build a set of candidate objects by excluding those objects that are less likely to correspond to traffic lights. Filtering may be based on various characteristics associated with traffic lights, such as shape, dimensions, texture, and location (for example, relative to vehicle 200). Such characteristics may be based on multiple examples of traffic lights and traffic control signals and may be stored in a database. In some embodiments, processing unit 110 may perform multi-frame analysis on the set of candidate objects that reflect possible traffic lights. For example, the processing unit 110 may track candidate objects across consecutive image frames, estimate their real-world locations, and remove those moving objects (which are less likely to be traffic lights). In some embodiments, the processing unit 110 may perform color analysis on the candidate objects to identify the relative positions of detected colors appearing within possible traffic lights.

[0177] In step 562, the processing unit 110 may analyze the geometric shape of the junction. The analysis may be based on any combination of (i) the number of lanes detected on any side of the vehicle 200, (ii) markings detected on the road (e.g., arrow marks), and (iii) a description of the junction extracted from map data (e.g., data from the map database 160). The processing unit 110 may perform the analysis using information derived from the execution of the monocular analysis module 402. Furthermore, the processing unit 110 may determine the correspondence between the traffic signals detected in step 560 and the lanes that appear near the vehicle 200.

[0178] As the vehicle 200 approaches the junction, in step 564, the processing unit 110 may update the confidence level associated with the analyzed junction geometry and detected traffic lights. For example, the number of traffic lights estimated to appear at the junction, compared to the number actually appearing at the junction, may influence the confidence level. Based on the confidence level, the processing unit 110 may delegate control to the driver of the vehicle 200 to improve safety conditions. By performing steps 560, 562, and 564, the processing unit 110 may identify the traffic lights appearing in the acquired set of images and analyze the geometric information of the junction. Based on the identification and analysis, the processing unit 110 may provide one or more navigation responses in the vehicle 200, as described in relation to Figure 5A above.

[0179] Figure 5E is a flowchart illustrating an exemplary process 500E for providing one or more navigation responses in a vehicle 200 based on a vehicle path, consistent with the disclosed embodiments. In step 570, the processing unit 110 may construct an initial vehicle path associated with the vehicle 200. The vehicle path may be represented using a set of points represented in coordinates (x,z), where d is the distance between two points in the set of points. i This can be within a range of 1 to 5 meters. In one embodiment, the processing unit 110 may construct an initial vehicle path using two polynomials, for example, left and right road polynomials. The processing unit 110 may calculate the geometric midpoint between the two polynomials and offset each point in the resulting vehicle path by a predetermined offset (e.g., smart lane offset), if any (an offset of zero may correspond to driving in the center of the lane). The offset may be made perpendicular to the segment between any two points in the vehicle path. In another embodiment, the processing unit 110 may use one polynomial and the estimated lane width to offset each point in the vehicle path by half the estimated lane width plus a predetermined offset (e.g., smart lane offset).

[0180] In step 572, the processing unit 110 may update the vehicle route constructed in step 570. The processing unit 110 calculates the distance d between two points in the set of points representing the vehicle route. k The distance d mentioned above i To make it shorter, the vehicle route constructed in step 570 may be reconstructed using a higher resolution. For example, distance d k This can fall within the range of 0.1 to 0.3 meters. The processing unit 110 may reconstruct the vehicle path using a parabolic spline algorithm that can generate a cumulative distance vector S corresponding to the total length of the vehicle path (i.e., based on a set of points representing the vehicle path).

[0181] In stage 574, the processing unit 110, based on the updated vehicle route constructed in stage 572, ((x l ,z l The system may determine look-ahead points (represented in coordinates as ). The processing unit 110 may extract look-ahead points from the cumulative distance vector S, and the look-ahead points may be associated with the look-ahead distance and look-ahead time. The look-ahead distance, which may have a lower limit in the range of 10 to 20 meters, can be calculated as the product of the vehicle speed 200 and the look-ahead time. For example, if the speed of the vehicle 200 decreases, the look-ahead distance may also decrease (for example, until it reaches the lower limit). Conversely, the look-ahead time, which may be in the range of 0.5 to 1.5 seconds, may be proportional to the gain of one or more control loops associated with providing the navigation response in the vehicle 200, such as the error tracking control loop in the direction of travel. For example, the gain of the error tracking control loop in the direction of travel may depend on the bandwidth of the yaw rate loop, the steering actuator loop, and the lateral motion state of the vehicle. Therefore, as the gain of the error tracking control loop in the direction of travel increases, the look-ahead time decreases.

[0182] In step 576, the processing unit 110 may determine the error in the direction of travel and the yaw rate command based on the lookahead point determined in step 574. The processing unit 110 may determine the arctangent of the lookahead point, for example, arctan(xl / z l The error in the traveling direction may be determined by calculating ). The processing unit 110 may determine a yaw rate command as the product of the error in the traveling direction and the high-level control gain. When the look-ahead distance is not at the lower limit, the high-level control gain may be equal to (2 / look-ahead time). Otherwise, the high-level control gain may be equal to (2 × the speed of the vehicle 200 / look-ahead distance).

[0183] FIG. 5F is a flowchart showing an exemplary process 500F for determining whether a preceding vehicle is changing lanes, which is not inconsistent with the disclosed embodiment. At step 580, the processing unit 110 may determine the navigation information associated with the preceding vehicle (e.g., the vehicle traveling ahead of the vehicle 200). For example, the processing unit 110 may determine the position, speed (e.g., direction and speed), and / or acceleration of the preceding vehicle using the techniques described in relation to FIGS. 5A and 5B above. The processing unit 110 may determine one or more road polynomials, the look-ahead points (associated with the vehicle 200), and / or the snail trail (e.g., a set of points describing the path taken by the preceding vehicle) using the techniques described in relation to FIG. 5E above.

[0184] In step 582, the processing unit 110 may analyze the navigation information determined in step 580. In one embodiment, the processing unit 110 may calculate the distance (e.g., along the trajectory) between the snail trail and the road polynomial. If the variance of this distance along the trajectory exceeds a predetermined threshold (e.g., 0.1 to 0.2 meters for straight roads, 0.3 to 0.4 meters for moderately curved roads, and 0.5 to 0.6 meters for roads with sharp curves), the processing unit 110 may determine that the preceding vehicle is likely to change lanes. If it is detected that multiple vehicles are traveling ahead of vehicle 200, the processing unit 110 may compare the snail trails associated with each vehicle. Based on the comparison, the processing unit 110 may determine that a vehicle whose snail trail does not match the snail trails of other vehicles is likely to change lanes. The processing unit 110 may further compare the curvature of the snail trail (associated with the preceding vehicle) with the expected curvature of the road segment on which the preceding vehicle is traveling. The expected curvature may be extracted from map data (for example, data from the map database 160), road polynomials, snail trails of other vehicles, and prior knowledge about the road. If the difference between the curvature of the snail trail and the expected curvature of the road segment exceeds a predetermined threshold, the processing unit 110 may determine that the preceding vehicle is likely to change lanes.

[0185] In another embodiment, the processing unit 110 may compare the instantaneous position of the preceding vehicle with a look-ahead point (associated with the vehicle 200) over a specific period (e.g., 0.5 to 1.5 seconds). If the distance between the instantaneous position of the preceding vehicle and the look-ahead point changes during the specific period, and the cumulative sum of the changes exceeds a predetermined threshold (e.g., 0.3 to 0.4 meters on a straight road, 0.7 to 0.8 meters on a moderately curved road, and 1.3 to 1.7 meters on a road with sharp curves), the processing unit 110 may determine that the preceding vehicle is likely to change lanes. In another embodiment, the processing unit 110 may analyze the geometry of the snail trail by comparing the lateral distance traveled along the trajectory with the expected curvature of the snail trail. The expected radius of curvature is calculated (δ z 2 +δ x 2 ) / 2 / (δ x The determination may be made according to δ, where δ x δ represents the lateral distance traveled, z This represents the longitudinal distance traveled. If the difference between the lateral distance traveled and the expected curvature exceeds a predetermined threshold (e.g., 500 to 700 meters), the processing unit 110 may determine that the preceding vehicle is likely to change lanes. In another embodiment, the processing unit 110 may analyze the position of the preceding vehicle. If the position of the preceding vehicle makes the road polynomial (e.g., the preceding vehicle is overlaid on top of the road polynomial) invisible, the processing unit 110 may determine that the preceding vehicle is likely to change lanes. If the preceding vehicle is positioned such that another vehicle is detected in front of it and the snail trails of the two vehicles are not parallel, the processing unit 110 may determine that the (closer) preceding vehicle is likely to change lanes.

[0186] In step 584, the processing unit 110 may determine whether the preceding vehicle 200 is changing lanes based on the analysis performed in step 582. For example, the processing unit 110 may make a determination based on a weighted average of the individual analyses performed in step 582. Under such a scheme, for example, the processing unit 110's determination that the preceding vehicle is likely to change lanes based on a particular type of analysis may be assigned a value of "1" (where "0" represents a determination that the preceding vehicle is unlikely to change lanes). Different analyses performed in step 582 may be assigned different weights, and the disclosed embodiments are not limited to any particular combination of analyses and weights.

[0187] Figure 6 is a flowchart illustrating an exemplary process 600 for performing one or more navigation responses based on stereo image analysis, consistent with the disclosed embodiments. In step 610, the processing unit 110 may receive a plurality of first and second images via the data interface 128. For example, cameras included in the image acquisition unit 120 (e.g., image acquisition devices 122 and 124 having fields of view 202 and 204) may acquire a plurality of first and second images of the area in front of the vehicle 200 and transmit them to the processing unit 110 via a digital connection (e.g., USB, wireless, Bluetooth®, etc.). In some embodiments, the processing unit 110 may receive a plurality of first and second images via two or more data interfaces. The disclosed embodiments are not limited to a specific data interface configuration or protocol.

[0188] In step 620, the processing unit 110 executes the stereo image analysis module 404 and performs stereo image analysis on the first and second sets of images to create a 3D map of the road ahead of the vehicle and may detect features in the images, such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, and road obstacles. The stereo image analysis may be performed in a manner similar to the steps described in relation to Figures 5A to 5D above. For example, the processing unit 110 may execute the stereo image analysis module 404 to detect candidate objects (e.g., vehicles, pedestrians, road markings, traffic lights, road obstacles, etc.) in the first and second sets of images, remove a subset of candidate objects based on various criteria, perform multi-frame analysis to construct measurements, and determine the confidence level for the remaining candidate objects. When performing the above steps, the processing unit 110 may consider information from both the first and second sets of images rather than information from only one set of images. For example, the processing unit 110 may analyze the differences in pixel-level data (or other data subsets from the two streams of captured images) for candidate objects that appear in both the first and second sets of images. As another example, the processing unit 110 may estimate the position and / or velocity of a candidate object (e.g., relative to the vehicle 200) by observing that an object appears in one of the sets of images but not in the other, or in relation to other differences that may exist for objects that appear in the two image streams. For example, the position, velocity and / or acceleration relative to the vehicle 200 may be determined based on the trajectory, position, motion characteristics, etc., of features associated with the object that appears in one or both of the image streams.

[0189] In step 630, the processing unit 110 may execute the navigation response module 408 to perform one or more navigation responses in the vehicle 200 based on the analysis performed in step 620 and the techniques described above in relation to Figure 4. Navigation responses may include, for example, changes in direction, lane changes, changes in acceleration, changes in speed, and braking. In some embodiments, the processing unit 110 may perform one or more navigation responses using data derived from the execution of the speed and acceleration module 406. Furthermore, multiple navigation responses may occur simultaneously, sequentially, or in any combination thereof.

[0190] Figure 7 is a flowchart illustrating an exemplary process 700 for making one or more navigation responses based on the analysis of three sets of images, consistent with the disclosed embodiments. In step 710, the processing unit 110 may receive first, second, and third sets of images via the data interface 128. For example, cameras included in the image acquisition unit 120 (e.g., image acquisition devices 122, 124, and 126 having fields of view 202, 204, and 206) may capture first, second, and third sets of images of the front and / or lateral regions of the vehicle 200 and transmit them to the processing unit 110 via a digital connection (e.g., USB, wireless, Bluetooth®, etc.). In some embodiments, the processing unit 110 may receive first, second, and third sets of images via three or more data interfaces. For example, each of the image acquisition devices 122, 124, and 126 may have a data interface associated with communicating data to the processing unit 110. The disclosed embodiments are not limited to any particular data interface configuration or protocol.

[0191] In step 720, the processing unit 110 may analyze the first, second, and third plurality of images to detect features in the images, such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic signals, and road obstacles. The analysis may be performed in a manner similar to the steps described in relation to Figures 5A to 5D and Figure 6 above. For example, the processing unit 110 may perform monocular image analysis on each of the first, second, and third plurality of images (for example, via the execution of the monocular image analysis module 402, based on the steps described in relation to Figures 5A to 5D above). Alternatively, the processing unit 110 may perform stereo image analysis on the first and second plurality of images, the second and third plurality of images, and / or the first and third plurality of images (for example, via the execution of the stereo image analysis module 404, based on the steps described in relation to Figure 6 above). The processed information corresponding to the analysis of the first, second, and / or third plurality of images may be combined. In some embodiments, the processing unit 110 may perform a combination of monocular and stereo image analysis. For example, the processing unit 110 may perform monocular image analysis on a first set of images (e.g., via the execution of the monocular image analysis module 402) and perform stereo image analysis on second and third sets of images (e.g., via the execution of the stereo image analysis module 404). The configuration of the image acquisition devices 122, 124, and 126 (including their respective positions and fields of view 202, 204, and 206) may affect the type of analysis performed on the first, second, and third sets of images. The disclosed embodiments are not limited to the specific configuration of the image acquisition devices 122, 124, and 126, or the type of analysis performed on the first, second, and third sets of images.

[0192] In some embodiments, the processing unit 110 may perform tests on the system 100 based on the images acquired and analyzed in steps 710 and 720. Such tests may provide an indicator of the overall performance of the system 100 for specific configurations of the image acquisition devices 122, 124, and 126. For example, the processing unit 110 may determine the percentage of "false hits" (e.g., when the system 100 incorrectly identifies the presence of a vehicle or pedestrian) and "misses."

[0193] In step 730, the processing unit 110 may provide one or more navigation responses for the vehicle 200 based on information derived from two of the first, second, and third images. The selection of two of the first, second, and third images may depend on various factors, such as the number, type, and size of objects detected in each of the images. The processing unit 110 may also make selections based on image quality and resolution, the effective field of view reflected in the images, the number of captured frames, and the extent to which one or more objects of interest actually appear in the frames (e.g., the percentage of frames in which objects appear, the proportion of objects appearing in each such frame, etc.).

[0194] In some embodiments, the processing unit 110 may select information derived from two of the first, second, and third images by determining the extent to which the information derived from one image source is consistent with the information derived from other image sources. For example, the processing unit 110 may combine the processed information derived from each of the image acquisition devices 122, 124, and 126 (whether from monocular analysis, stereo analysis, or any combination of the two) to determine visual indicators (e.g., lane markings, detected vehicles and their positions and / or routes, detected traffic lights, etc.) that match across the images acquired from each of the image acquisition devices 122, 124, and 126. The processing unit 110 may exclude inconsistent information across the acquired images (e.g., vehicles changing lanes, lane models indicating vehicles too close to vehicle 200, etc.). Therefore, the processing unit 110 may select information derived from two of the first, second, and third images based on its determination of consistent and inconsistent information.

[0195] Navigation responses may include, for example, changes in direction, lane changes, and changes in acceleration. Processing unit 110 may perform one or more navigation responses based on the analysis performed in step 720 and the techniques described above in relation to Figure 4. Processing unit 110 may perform one or more navigation responses using data derived from the performance of the velocity and acceleration module 406. In some embodiments, processing unit 110 may perform one or more navigation responses based on the relative position, relative velocity, and / or relative acceleration between an object detected in any of the first, second, and third images and the vehicle 200. Multiple navigation responses may occur simultaneously, sequentially, or in any combination thereof.

[0196] Sparse road models for autonomous vehicle navigation

[0197] In some embodiments, the disclosed systems and methods may use sparse maps for the navigation of an autonomous vehicle. In particular, the sparse map may be for the navigation of an autonomous vehicle along a road segment. For example, the sparse map may provide sufficient information to navigate an autonomous vehicle without storing and / or updating large amounts of data. As will be discussed in more detail below, an autonomous vehicle may use a sparse map to navigate one or more roads based on one or more stored trajectories.

[0198] Sparse map for autonomous vehicle navigation

[0199] In some embodiments, the disclosed systems and methods may generate sparse maps for the navigation of an autonomous vehicle. For example, a sparse map may provide sufficient information for navigation without requiring excessive data storage or data transfer rates. As will be discussed in more detail below, a vehicle (which may be an autonomous vehicle) may use a sparse map to navigate one or more roads. For example, in some embodiments, a sparse map may include data relating to roads and potential landmarks along those roads, which may be sufficient for vehicle navigation but also have a small data footprint. For example, sparse data maps, as described in detail below, may require significantly less storage space and data transfer bandwidth compared to digital maps that include detailed map information, such as image data collected along roads.

[0200] For example, rather than storing detailed representations of road segments, a sparse data map may store a three-dimensional polynomial representation of preferred vehicle paths along the road. These paths require little to no data storage space. Furthermore, in the sparse data map described, landmarks may be identified and included in the sparse map road model for use in navigation. These landmarks may be placed at any intervals appropriate to enable vehicle navigation, but in some cases, such landmarks do not need to be identified and included in the model at high density and short intervals. Rather, in some cases, navigation may be possible based on landmarks spaced at intervals of at least 50 meters, at least 100 meters, at least 500 meters, at least 1 kilometer, or at least 2 kilometers. As will be discussed in more detail in other sections, a sparse map may be generated based on data collected or measured by a vehicle equipped with various sensors and devices, such as image acquisition devices, global positioning system sensors, motion sensors, etc., as the vehicle travels along the roadway. In some cases, a sparse map may be generated based on data collected during multiple drives of one or more vehicles along a particular roadway. Generating a sparse map using multiple drives of one or more vehicles may be referred to as "crowdsourcing" the sparse map.

[0201] Without regard to the disclosed embodiments, an autonomous vehicle system may use a sparse map for navigation. For example, the disclosed system and method may supply a sparse map to generate a road navigation model for an autonomous vehicle, and may use the sparse map and / or the generated road navigation model to navigate the autonomous vehicle along a road segment. A sparse map not inconsistent with the disclosure may include one or more three-dimensional contours that can represent predetermined trajectories that the autonomous vehicle may traverse as it moves along the relevant road segment.

[0202] A sparse map not inconsistent with this disclosure may include data representing one or more road features. Such road features may include recognized landmarks, road signature profiles, and any other road-related features useful for navigating a vehicle. A sparse map not inconsistent with this disclosure may enable autonomous navigation of a vehicle based on a relatively small amount of data contained in the sparse map. For example, rather than including data detailing a detailed representation of a road, such as road edges, road curvature, images associated with road segments, or other physical features associated with road segments, the disclosed embodiments of a sparse map may require relatively small storage space (and, if any portion of the sparse map is transferred to a vehicle, relatively small bandwidth), but may still be sufficient for the navigation of an autonomous vehicle. The small data footprint of the disclosed sparse maps, discussed in more detail below, can be achieved in some embodiments by storing representations of road-related elements that require only small amounts of data, while still enabling autonomous navigation.

[0203] For example, rather than storing detailed representations of various aspects of a road, the disclosed sparse map may store polynomial representations of one or more trajectories a vehicle may take along the road. Therefore, rather than needing to store (or transfer) details about the physical properties of the road to enable navigation along a road using the disclosed sparse map, the vehicle may be navigated along a particular road segment by aligning its travel path with a trajectory (e.g., a polynomial spline) along that particular road segment, without needing to interpret the physical aspects of the road in some cases. In this way, the vehicle may be navigated primarily on stored trajectories (e.g., polynomial splines), which may require considerably less storage space than methods that include storing roadway images, road parameters, road layouts, etc.

[0204] In addition to the polynomial representation of the stored trajectory along the road segment, the disclosed sparse map may include small data objects that can represent road features. In some embodiments, the small data objects may include a digital signature, which is derived from a digital image (or digital signal) acquired from a sensor (e.g., a camera or other sensor such as a suspension sensor) mounted on a vehicle traveling along the road segment. The digital signature may have a reduced size relative to the signal acquired by the sensor. In some embodiments, the digital signature may be constructed to be compatible with a classification function configured, for example, to detect and identify road features from signals acquired by the sensor during subsequent driving. In some embodiments, at a later point in time, based on an image of the road feature (or, if the stored signature is not image-based and / or includes other data, a digital signal generated by a sensor) of the road feature captured by a camera-equipped vehicle traveling along the same road segment, the digital signature may be constructed such that it has as small a footprint as possible, while maintaining the ability for the road feature to correlate with or match the stored signature.

[0205] In some embodiments, the size of the data object may further be associated with the uniqueness of the road feature. For example, for road features detectable by a camera mounted on a vehicle, the camera system mounted on the vehicle may be coupled to a classifier capable of distinguishing image data corresponding to a particular type of road feature, such as a road sign, where such a road sign is locally unique in its area (e.g., no identical or similar road signs are nearby) and the data may be sufficient to store data indicating the type of road feature and its location.

[0206] As will be discussed in more detail below, road features (e.g., landmarks along a road segment) may be stored as small data objects that can represent the road features in a relative number of bytes, while simultaneously providing sufficient information to recognize and use such features for navigation. For example, road signs may be identified as recognized landmarks that can form the basis of vehicle navigation. The representation of a road sign may be stored in a sparse map containing, for example, a few bytes of data indicating the type of landmark (e.g., a stop sign) and a few bytes of data indicating the location of the landmark (e.g., coordinates). Navigation based on such a light data representation of a landmark (e.g., using a representation sufficient to search, recognize, and navigate based on the landmark) may provide the desired level of navigation functionality associated with a sparse map without significantly increasing the data overhead associated with the sparse map. This lean representation of landmarks (and other road features) may utilize sensors and processors contained in and mounted on such a vehicle that are configured to detect, identify, and / or classify specific road features.

[0207] For example, if a sign, or more specifically, a particular type of sign, is locally unique in a given region (e.g., no other signs exist, or no other signs of the same type exist), the sparse map may use data indicating the type of landmark (sign, or particular type of sign). During navigation (e.g., autonomous navigation), if a camera mounted on an autonomous vehicle captures an image of a region containing a sign (or a particular type of sign), the processor may process the image, detect the sign (if it actually exists in the image), classify the image as a sign (or a particular type of sign), and correlate the location of the image with the location of the sign stored in the sparse map.

[0208] Generating a sparse map

[0209] In some embodiments, the sparse map may include line representations of road surface features extending along a road segment and at least one of a plurality of landmarks associated with the road segment. In certain embodiments, the sparse map may be generated via "crowdsourcing," for example, through image analysis of a plurality of images taken as one or more vehicles pass through the road segment.

[0210] Figure 8 shows a sparse map 800 that one or more vehicles, for example, vehicle 200 (which may be an autonomous vehicle), can access to provide navigation for the autonomous vehicle. The sparse map 800 may be stored in memory, for example, memory 140 or 150. Such memory devices may include any type of non-temporary storage device or computer-readable medium. For example, in some embodiments, memory 140 or 150 may include a hard drive, compact disk, flash memory, magnetic-based memory device, optical-based memory device, etc. In some embodiments, the sparse map 800 may be stored in a database (e.g., map database 160) which may be stored in memory 140 or 150 or other types of storage devices.

[0211] In some embodiments, the sparse map 800 may be stored in a storage device or non-transient computer-readable medium provided on board the vehicle 200 (for example, a storage device included in a navigation system mounted on the vehicle 200). A processor provided to the vehicle 200 (for example, a processing unit 110) may access the sparse map 800 stored in the storage device or computer-readable medium provided on board the vehicle 200 to generate navigation commands for guiding the autonomous vehicle 200 as the vehicle passes through a road segment.

[0212] However, the sparse map 800 does not need to be stored locally with respect to the vehicle. In some embodiments, the sparse map 800 may be stored in a storage device or computer-readable medium provided to a remote server that communicates with the vehicle 200 or a device associated with the vehicle 200. A processor provided to the vehicle 200 (e.g., processing unit 110) may receive data contained in the sparse map 800 from the remote server and execute data to guide the autonomous driving of the vehicle 200. In such embodiments, the remote server may store all or only a portion of the sparse map 800. Thus, a storage device or computer-readable medium provided mounted on the vehicle 200 and / or one or more additional vehicles may store the remaining portion of the sparse map 800.

[0213] Furthermore, in such embodiments, the sparse map 800 may be accessible to multiple vehicles traversing various road segments (e.g., tens, hundreds, thousands, or millions of vehicles). It should also be noted that the sparse map 800 may include multiple submaps. For example, in some embodiments, the sparse map 800 may include hundreds, thousands, hundreds, or more submaps that can be used when navigating a vehicle. Such submaps may be referred to as local maps, and a vehicle traveling along a roadway may access any number of local maps relevant to the vehicle's location. The local map portion of the sparse map 800 may be stored using a Global Navigation Satellite System (GNSS) key as an index to the database of the sparse map 800. Thus, in this system, the calculation of steering angles for navigating a host vehicle can be performed without relying on the host vehicle's GNSS position, road features, or landmarks, while such GNSS information may be used for retrieving relevant local maps.

[0214] Generally, the sparse map 800 may be generated based on data collected from one or more vehicles as they travel along a roadway. For example, sensors mounted on one or more vehicles (e.g., cameras, speedometers, GPS, accelerometers, etc.) may be used to record the trajectories of one or more vehicles traveling along a roadway, and a polynomial representation of a preferred trajectory for a vehicle following along the roadway may be determined based on the collected trajectories traveled by one or more vehicles. Similarly, data collected by one or more vehicles may be useful in identifying potential landmarks along a particular roadway. Data collected from passing vehicles may be used to identify road profile information, such as road width profiles, road roughness profiles, traffic spacing profiles, and road conditions. Using the collected information, the sparse map 800 may be generated and supplied (e.g., to local storage or via on-the-fly data transmission) for use when navigating one or more autonomous vehicles. However, in some embodiments, map generation may not be completed at the time of initial map generation. As will be discussed in more detail below, while those vehicles continue to travel along the roadways included in sparse map 800, sparse map 800 may be continuously or periodically updated based on data collected from the vehicles.

[0215] The data recorded in the sparse map 800 may include location information based on Global Positioning System (GPS) data. For example, location information may be included in the sparse map 800 for various map elements, including the locations of landmarks and road profiles. The locations of map elements included in the sparse map 800 may be obtained using GPS data collected from vehicles traveling on the roadway. For example, a vehicle passing a identified landmark may determine the location of the identified landmark using GPS location information associated with the vehicle and a determination of the location of the identified landmark relative to the vehicle (for example, based on image analysis of data collected from one or more cameras mounted on the vehicle). Such a location determination of the identified landmark (or any other feature included in the sparse map 800) may be repeated when additional vehicles pass over the location of the identified landmark. Some or all of the additional location determinations may be used to refine the location information stored in the sparse map 800 for the identified landmark. For example, in some embodiments, multiple location measurements for a particular feature stored in the sparse map 800 may be averaged together. However, any other mathematical operation may be used to refine the stored location of a map element based on multiple determined locations for that map element.

[0216] The sparse maps of the disclosed embodiments may enable autonomous vehicle navigation using a relatively small amount of stored data. In some embodiments, the sparse map 800 may have a data density of less than 2 MB per kilometer of road, less than 1 MB per kilometer of road, less than 500 kB per kilometer of road, or less than 100 kB per kilometer of road (e.g., including data representing target trajectories, landmarks, and any other stored road features). In some embodiments, the data density of the sparse map 800 may be less than 10 kB per kilometer of road, more specifically less than 2 kB per kilometer of road (e.g., 1.6 kB per kilometer), or 10 kB or less per kilometer of road, or 20 kB or less per kilometer of road. In some embodiments, most, if not all, of the roadways in the United States may be autonomously navigated using a sparse map with a total data of 4 GB or less. These data density values ​​may represent the average across the entire sparse map 800, across local maps within sparse map 800, and / or across specific road segments within sparse map 800.

[0217] As described above, the sparse map 800 may include representations of multiple target trajectories 810 for guiding autonomous driving or navigation along road segments. Such target trajectories may be stored as three-dimensional splines. For example, target trajectories stored in the sparse map 800 may be determined based on two or more reconstructed trajectories from a vehicle's previous passage along a particular road segment. A road segment may be associated with a single target trajectory or multiple target trajectories. For example, on a two-lane road, a first target trajectory may be stored to represent a target route for travel along the road in a first direction, and a second target trajectory may be stored to represent a target route for travel along the road in another direction (e.g., opposite to the first direction). Additional target trajectories may be stored for a particular road segment. For example, on a multi-lane road, one or more target trajectories may be stored representing target routes for vehicles in one or more lanes associated with the multi-lane road. In some embodiments, each lane of a multi-lane road may be associated with its own target trajectory. In other embodiments, fewer target trajectories may be stored than the number of lanes present on a multi-lane road. In such cases, a vehicle navigating a multi-lane road may use any of the stored target trajectories to guide its navigation, taking into account the amount of lane offset from the lane in which the target trajectory is stored (for example, if a vehicle is traveling in the leftmost lane of a three-lane highway and target trajectories are stored only for the center lane of the highway, the vehicle may use the target trajectory for the center lane to navigate by taking into account the amount of lane offset between the center lane and the leftmost lane when generating navigation commands).

[0218] In some embodiments, the target trajectory may represent the ideal path that a vehicle should take when traveling. The target trajectory may be positioned, for example, approximately in the center of the driving lane. In other cases, the target trajectory may be positioned at other locations relative to the road segment. For example, the target trajectory may substantially coincide with the center of the road, the edge of the road, or the edge of the lane. In such cases, navigation based on the target trajectory may include a determined amount of offset to be maintained relative to the position of the target trajectory. Furthermore, in some embodiments, the determined amount of offset to be maintained relative to the position of the target trajectory may differ based on the type of vehicle (for example, a passenger car with two axles along at least part of the target trajectory may have a different offset than a truck with more than two axles).

[0219] The sparse map 800 may include data related to a plurality of predetermined landmarks 820 associated with specific road segments, local maps, etc. These landmarks can be used during the navigation of an autonomous vehicle, as will be discussed in more detail below. For example, in some embodiments, landmarks can be used to determine the vehicle's current position relative to a stored target trajectory. Using this position information, the autonomous vehicle can adjust its direction of travel to match the direction of the target trajectory at the determined position.

[0220] Multiple landmarks 820 may be identified and stored in the sparse map 800 at any appropriate intervals. In some embodiments, landmarks may be stored at a relatively high density (e.g., every few meters or more). However, in some embodiments, fairly large landmark spacing values ​​may be used. For example, in the sparse map 800, identified (or recognized) landmarks may be spaced 10 meters, 20 meters, 50 meters, 100 meters, 1 kilometer, or 2 kilometers apart. In some cases, identified landmarks may be located at distances greater than 2 kilometers.

[0221] Between landmarks, and therefore during the determination of the vehicle's position relative to the target trajectory, the vehicle may navigate based on dead reckoning, in which the vehicle uses sensors to determine its egomotion and estimate its position relative to the target trajectory. Since errors can accumulate during navigation by dead reckoning, the determination of the position over time relative to the target trajectory may become increasingly inaccurate. The vehicle may use landmarks present in the sparse map 800 (their known positions) to eliminate the errors induced by dead reckoning in position determination. In this way, identified landmarks included in the sparse map 800 can serve as navigation anchors from which the vehicle's precise position relative to the target trajectory can be determined. Identified landmarks do not need to always be available to the autonomous vehicle, as a certain amount of error may be acceptable at appropriate locations. Rather, appropriate navigation may still be based on landmark spacing of 10 meters, 20 meters, 50 meters, 100 meters, 500 meters, 1 kilometer, 2 kilometers, or longer, as described above. In some embodiments, a density of one identified landmark per kilometer of road may be sufficient to maintain longitudinal positioning accuracy within 1 meter. Therefore, it is not necessary for all possible landmarks appearing along a road segment to be stored in the sparse map 800.

[0222] Furthermore, in some embodiments, lane markings may be used to locate the vehicle's position during landmark spacing. By using lane markings during landmark spacing, accumulation during navigation due to dead reckoning can be minimized.

[0223] In addition to the target trajectory and identified landmarks, the sparse map 800 may include information related to various other road features. For example, Figure 9A shows a representation of a curve along a particular road segment that may be stored in the sparse map 800. In some embodiments, a single lane of a road may be modeled by a three-dimensional polynomial description of the left and right sides of the road. Such polynomials representing the left and right sides of a single lane are shown in Figure 9A. Regardless of how many lanes a road may have, the road may be represented using polynomials in a similar manner to that shown in Figure 9A. For example, the left and right sides of a multi-lane road may be represented by polynomials similar to those shown in Figure 9A, and intermediate lane markings included in a multi-lane road (e.g., dashed markings representing lane boundaries, solid yellow lines representing boundaries between lanes traveling in different directions, etc.) may be represented using polynomials such as those shown in Figure 9A.

[0224] As shown in Figure 9A, lane 900 may be represented using a polynomial (e.g., a first-order, second-order, third-order, or any appropriate degree polynomial). For example, lane 900 is shown as a two-dimensional lane, and the polynomial is shown as a two-dimensional polynomial. As shown in Figure 9A, lane 900 includes left lane 910 and right lane 920. In some embodiments, polynomials greater than 1 may be used to represent the position on each side of the road or lane boundary. For example, each of left lane 910 and right lane 920 may be represented by multiple polynomials of any appropriate length. In some cases, the polynomials may have a length of about 100m, but other lengths longer or shorter than 100m may be used. Furthermore, the polynomials may overlap each other to facilitate a seamless transition when navigating based on the polynomials encountered later as the host vehicle travels along the roadway. For example, each of the left 910 and right 920 may be separated into segments of approximately 100 meters in length (an example of a predetermined first range) and represented by multiple cubic polynomials that overlap each other for approximately 50 meters. The polynomials representing the left 910 and right 920 may or may not have the same degree. For example, in some embodiments, some polynomials may be quadratic, some may be cubic, and some may be quartic.

[0225] In the example shown in Figure 9A, the left side 910 of lane 900 is represented by two groups of cubic polynomials. The first group includes polynomial segments 911, 912, and 913. The second group includes polynomial segments 914, 915, and 916. The two groups are substantially parallel to each other, while following their respective positions on the road. Polynomial segments 911, 912, 913, 914, 915, and 916 have a length of approximately 100 meters and overlap by approximately 50 meters with continuously adjacent segments. However, as mentioned above, polynomials of different lengths and different amounts of overlap may be used. For example, the polynomials may have lengths of 500m, 1km, or longer, and the amount of overlap may vary from 0 to 50m, from 50m to 100m, or greater than 100m. Furthermore, although Figure 9A is shown as representing polynomials extended to 2D space (e.g., on the surface of paper), it should be understood that these polynomials can represent curves that extend to three dimensions (e.g., including a height component) to represent elevation changes within road segments in addition to XY curvature. In the example shown in Figure 9A, the right side 920 of lane 900 is further represented by a first group having polynomial segments 921, 922, and 923, and a second group having polynomial segments 924, 925, and 926.

[0226] Returning to the target trajectories in sparse map 800, Figure 9B shows a three-dimensional polynomial representing the target trajectory for a vehicle traveling along a specific road segment. The target trajectory represents not only the XY path that the host vehicle should travel along a specific road segment, but also the elevation changes the host vehicle experiences as it travels along the road segment. Thus, each target trajectory in sparse map 800 can be represented by one or more three-dimensional polynomials, such as the three-dimensional polynomial 950 shown in Figure 9B. Sparse map 800 may contain multiple trajectories (for example, millions or billions or more trajectories to represent the trajectories of vehicles along various road segments along roadways around the world). In some embodiments, each target trajectory may correspond to a spline connecting three-dimensional polynomial segments.

[0227] Regarding the data footprint of the polynomial curves stored in the sparse map 800, in some embodiments, each cubic polynomial may be represented by four parameters, each requiring 4 bytes of data. A suitable representation can be obtained with a cubic polynomial requiring approximately 192 bytes of data per 100m. This may correspond to approximately 200kB per hour for data usage / transfer requirements for a host vehicle traveling at approximately 100km / h.

[0228] SparseMap800 can describe lane networks using a combination of geometric descriptors and metadata. As mentioned above, the geometric shape can be described by polynomials or splines. The metadata can describe the number of lanes, special characteristics (e.g., carpool lanes), and other possible sparse labels. The total footprint of such indicators can be very small.

[0229] Accordingly, the sparse map according to the embodiments disclosed herein may include at least one line representation of a road surface feature extending along a road segment, and each line representation representing a path along the road segment substantially corresponding to the road surface feature. As described above, in some embodiments, the at least one line representation of a road surface feature may include a spline, a polynomial representation, or a curve. Furthermore, in some embodiments, the road surface feature may include at least one of road edges or lane markings. In addition, as discussed below with respect to "crowdsourcing," the road surface feature may be identified through image analysis of multiple images acquired as one or more vehicles pass through the road segment.

[0230] As mentioned above, the sparse map 800 may include a number of predetermined landmarks associated with a road segment. Rather than storing real images of the landmarks and relying, for example, on image recognition analysis based on captured and stored images, each landmark in the sparse map 800 may be represented and recognized using less data than required by the stored real images. The data representing the landmarks may include more information sufficient to describe or identify the landmarks along the road. Storing data that describes the characteristics of the landmarks rather than the real images of the landmarks can reduce the size of the sparse map 800.

[0231] Figure 10 shows examples of types of landmarks that may be represented in the sparse map 800. Landmarks may include any visible and identifiable objects along a road segment. Landmarks may be selected so that they are constant and do not change much in terms of their location and / or content. Landmarks included in the sparse map 800 may be useful in determining the position of a vehicle 200 relative to a target trajectory as the vehicle passes through a particular road segment. Examples of landmarks may include traffic signs, directional signs, general signs (e.g., rectangular signs), roadside structures (e.g., lampposts, reflectors, etc.) and any other appropriate categories. In some embodiments, lane markings on the road may be included as landmarks in the sparse map 800.

[0232] Examples of landmarks shown in Figure 10 include traffic signs, directional signs, roadside structures, and general signs. Traffic signs may include, for example, speed limit signs (e.g., speed limit sign 1000), yield signs (e.g., yield sign 1005), route number signs (e.g., route number sign 1010), traffic signal light signs (e.g., traffic signal light sign 1015), and stop signs (e.g., stop sign 1020). Directional signs may include signs that include one or more arrows indicating one or more directions to different locations. For example, a directional sign may include a highway sign 1025 having an arrow for directing a vehicle to a different road or location, and an exit sign 1030 having an arrow for directing a vehicle to move away from the road. Thus, at least one of the multiple landmarks may include a road sign.

[0233] General signs may be unrelated to traffic. For example, general signs may include billboards used for advertising, or welcome signs adjacent to the boundary between two countries, states, counties, cities, or towns. Figure 10 shows a general sign 1040 ("Joe's Restaurant"). While a general sign 1040 may have a rectangular shape, as shown in Figure 10, a general sign 1040 may also have other shapes, such as a square, circle, or triangle.

[0234] Landmarks may include roadside fixtures. Roadside fixtures may be objects that are not signs and may not be related to traffic or direction. For example, roadside fixtures may include lampposts (e.g., lamppost 1035), utility poles, traffic signal poles, etc.

[0235] Landmarks may include beacons that can be specifically designed for use in the navigation systems of autonomous vehicles. For example, such beacons may include standalone structures that are placed at predetermined intervals to assist in navigating a host vehicle. Such beacons may include visual / graphic information (e.g., icons, emblems, barcodes, etc.) that can be added to existing road signs that can be identified or recognized by vehicles traveling along a road segment. Such beacons may include electronic components. In such embodiments, electronic beacons (e.g., RFID tags, etc.) may be used to transmit non-visual information to the host vehicle. Such information may include, for example, identification information of a landmark and / or location information of a landmark that can be used by the host vehicle to determine its position along a target trajectory.

[0236] In some embodiments, landmarks included in the sparse map 800 may be represented by data objects of predetermined size. The data representing a landmark may include any appropriate parameters for identifying a particular landmark. For example, in some embodiments, a landmark stored in the sparse map 800 may include parameters such as the physical size of the landmark (e.g., to support the estimation of distance to the landmark based on known size / scale), distance to a previous landmark, lateral offset, height, type code (e.g., type of landmark - any type of directional sign, traffic sign, etc.), GPS coordinates (e.g., to support global localization), and any other appropriate parameters. Each parameter may be associated with a data size. For example, the size of a landmark may be stored using 8 bytes of data. The distance to a previous landmark, lateral offset, and height may be defined using 12 bytes of data. A type code associated with a landmark such as a directional sign or traffic sign may require approximately 2 bytes of data. For general signs, an image signature enabling the identification of the general sign may be stored using 50 bytes of data storage. The GPS location of a landmark may be associated with 16 bytes of data storage. These data sizes for each parameter are merely examples; other data sizes may also be used.

[0237] Representing landmarks on the sparse map 800 in this manner may provide a lean solution for efficiently representing landmarks in a database. In some embodiments, signs may be referred to as semantic signs and non-semantic signs. Semantic signs may include signs of any classification with a standard meaning (e.g., speed limit signs, warning signs, directional signs, etc.). Non-semantic signs may include any signs not associated with a standard meaning (e.g., general advertising signs identifying businesses, etc.). For example, each semantic sign may be represented by 38 bytes of data (e.g., 8 bytes for size, 12 bytes for distance to previous landmark, lateral offset and height, 2 bytes for type code, and 16 bytes for GPS coordinates). The sparse map 800 may represent the type of landmark using a tagging system. In some cases, each traffic sign or directional sign may be associated with its own tag, which may be stored in the database as part of the landmark identification information. For example, the database may contain about 1,000 different tags to represent various traffic signs, and about 10,000 different tags to represent directional signs. Of course, any appropriate number of tags may be used, and additional tags may be created as needed. In some embodiments, a general-purpose sign may be represented using less than about 100 bytes (for example, about 86 bytes, including 8 bytes for size, 12 bytes for distance to previous landmark, lateral offset and height, 50 bytes for image signature, and 16 bytes for GPS coordinates).

[0238] Therefore, for semantic road signs that do not require image signatures, the data density impact on sparse map 800 can be approximately 760 bytes per kilometer (e.g., 20 landmarks per kilometer × 38 bytes per landmark = 760 bytes), even when the landmark density is relatively high at about one landmark per 50 meters. Even considering generic signs that include an image signature component, the data density impact is approximately 1.72 kB per kilometer (e.g., 20 landmarks per kilometer × 86 bytes per landmark = 1720 bytes). For semantic road signs, this corresponds to approximately 76 kB of data usage per hour for a vehicle traveling at 100 km / h. For generic signs, this corresponds to approximately 170 kB per hour for a vehicle traveling at 100 km / h.

[0239] In some embodiments, a generally rectangular object, such as a rectangular sign, may be represented in the sparse map 800 by data of 100 bytes or less. The representation of a generally rectangular object (e.g., a general sign 1040) in the sparse map 800 may include a simplified image signature (e.g., a simplified image signature 1045) associated with the generally rectangular object. This simplified image signature may be used, for example, to help identify a general sign as a recognized landmark. Such a simplified image signature (e.g., image information derived from real image data representing the object) can avoid the need to store the real image of the object or the need to compare it with image analysis performed on the real image to recognize the landmark.

[0240] Referring to Figure 10, the sparse map 800 may contain or store a simplified image signature 1045 associated with the general sign 1040, rather than the actual image of the general sign 1040. For example, an image acquisition device (e.g., image acquisition devices 122, 124, or 126) may acquire an image of the general sign 1040, and a processor (e.g., image processor 190, or any other processor capable of processing images, mounted on or located remotely from the host vehicle) may perform image analysis to extract / create a simplified image signature 1045 containing a unique signature or pattern associated with the general sign 1040. In one embodiment, the simplified image signature 1045 may include shape, color, pattern, brightness pattern, or any other features that can be extracted from an image of the general sign 1040 to describe the general sign 1040.

[0241] For example, in Figure 10, the circles, triangles, and stars shown in the simplified image signature 1045 may represent areas of different colors. The patterns represented by circles, triangles, and stars may be stored in the sparse map 800, for example, within 50 bytes designated to contain the image signature. In particular, the circles, triangles, and stars do not necessarily mean that such shapes are stored as part of the image signature. Rather, these shapes conceptually represent recognizable areas having identifiable color differences, texture areas, graphic shapes, or other variations of properties that can be associated with general-purpose signs. Such simplified image signatures may be used to identify landmarks in the form of general signs. For example, a simplified image signature may be used to perform a same-not-same analysis based on a comparison of image data captured using a camera mounted on an autonomous vehicle with a stored simplified image signature.

[0242] Thus, a plurality of landmarks can be identified through image analysis of a plurality of images acquired when one or more vehicles pass through a road segment. As will be described below with respect to "crowdsourcing", in some embodiments, image analysis for identifying a plurality of landmarks may include approving a potential landmark if the ratio of images in which the landmark appears to images in which the landmark does not appear exceeds a threshold. Further, in some embodiments, image analysis for identifying a plurality of landmarks may include rejecting a potential landmark if the ratio of images in which the landmark does not appear to images in which the landmark appears exceeds a threshold.

[0243] When the host vehicle returns to a target trajectory that can be used for navigation on a particular road segment, FIG. 11A shows a polynomial representation of the trajectory captured during the process of building or maintaining the sparse map 800. The polynomial representation of the target trajectory included in the sparse map 800 can be determined based on two or more reconstructed trajectories at the previous passage of the vehicle along the same road segment. In some embodiments, the polynomial representation of the target trajectory included in the sparse map 800 can be a set of two or more reconstructed trajectories at the previous passage of the vehicle along the same road segment. In some embodiments, the polynomial representation of the target trajectory included in the sparse map 800 can be an average of two or more reconstructed trajectories at the previous passage of the vehicle along the same road segment. Other mathematical operations can also be used to construct a target trajectory along a road route based on reconstructed trajectories collected from vehicles passing along the road segment.

[0244] As shown in FIG. 11A, a number of vehicles 200 may travel on a road segment 1100 at different times. Each vehicle 200 may collect data related to the route along which the vehicle has traveled on the road segment. The route traveled by a particular vehicle may be determined based on, among possible information sources, in particular camera data, accelerometer information, speed sensor information, and / or GPS information. Such data may be used to reconstruct the trajectories of vehicles traveling along the road segment, and based on these reconstructed trajectories, a target trajectory (or a plurality of target trajectories) may be determined for a particular road segment. Such a target trajectory may represent a preferred route of a host vehicle (e.g., guided by an autonomous navigation system) when the vehicle travels along the road segment.

[0245] In the example shown in FIG. 11A, a first reconstructed trajectory 1101 may be determined based on data received from a first vehicle passing through the road segment 1100 during a first time period (e.g., the first day), a second reconstructed trajectory 1102 may be obtained from a second vehicle passing through the road segment 1100 during a second time period (e.g., the second day), and a third reconstructed trajectory 1103 may be obtained from a third vehicle passing through the road segment 1100 during a third time period (e.g., the third day). Each of the trajectories 1101, 1102, and 1103 may be represented by a polynomial such as a three-dimensional polynomial. Note that in some embodiments, any of the reconstructed trajectories may be assembled on a vehicle passing through the road segment 1100.

[0246] Furthermore, or alternatively, such reconstructed trajectories may be determined on the server side based on information received from vehicles passing through the road segment 1100. For example, in some embodiments, a vehicle 200 may transmit data to one or more servers relating to its movements along the road segment 1100 (e.g., among many, steering angle, direction of travel, time, position, speed, detected road geometry, and / or detected landmarks). The servers may reconstruct trajectories for the vehicle 200 based on the received data. Based on the first, second, and third trajectories 1101, 1102, and 1103, the servers may generate target trajectories to guide the navigation of an autonomous vehicle subsequently traveling along the same road segment 1100. While a target trajectory may be associated with a single previous passage through the road segment, in some embodiments, each target trajectory included in the sparse map 800 may be determined based on two or more reconstructed trajectories of vehicles passing through the same road segment. In Figure 11A, the target trajectory is represented by 1110. In some embodiments, the target trajectory 1110 may be generated based on the average of the first, second, and third trajectories 1101, 1102, and 1103. In some embodiments, the target trajectory 1110 contained in the sparse map 800 may be a set of two or more reconstructed trajectories (e.g., a weighted combination).

[0247] Figures 11B and 11C further illustrate the concept of target trajectories associated with road segments located within a geographical area 1111. As shown in Figure 11B, a first road segment 1120 within the geographical area 1111 may include a multi-lane road with two lanes 1122 designated for vehicle travel in a first direction, and two additional lanes 1124 designated for vehicle travel in a second direction opposite to the first direction. Lanes 1122 and 1124 may be separated by double yellow lanes 1123. The geographical area 1111 may also include a branched road segment 1130 intersecting road segment 1120. Road segment 1130 may include a two-lane road, with each lane designated for a different direction of travel. The geographical area 1111 may also include other road features, such as stop lines 1132, stop signs 1134, speed limit signs 1136, and hazard signs 1138.

[0248] As shown in Figure 11C, the sparse map 800 may include a local map 1140 containing a road model to assist in the autonomous navigation of a vehicle within the geographical region 1111. For example, the local map 1140 may include target trajectories for one or more lanes associated with road segments 1120 and / or 1130 within the geographical region 1111. For example, the local map 1140 may include target trajectories 1141 and / or 1142 that an autonomous vehicle can access or rely on when passing through lane 1122. Similarly, the local map 1140 may include target trajectories 1143 and / or 1144 that an autonomous vehicle can access or rely on when passing through lane 1124. Furthermore, the local map 1140 may include target trajectories 1145 and / or 1146 that an autonomous vehicle can access or rely on when passing through road segment 1130. Target trajectory 1147 represents a preferred path that the autonomous vehicle should take when transitioning from lane 1120 (in particular, relating to target trajectory 1141 associated with the rightmost lane of lane 1120) to road segment 1130 (in particular, relating to target trajectory 1145 associated with the first side of road segment 1130). Similarly, target trajectory 1148 represents a preferred path that the autonomous vehicle should take when transitioning from road segment 1130 (in particular, relating to target trajectory 1146) to a portion of road segment 1124 (in particular, relating to target trajectory 1143 associated with the left lane of lane 1124, as shown).

[0249] The sparse map 800 may include representations of other road-related features associated with the geographical area 1111. For example, the sparse map 800 may include representations of one or more landmarks identified in the geographical area 1111. Such landmarks may include a first landmark 1150 associated with a stop line 1132, a second landmark 1152 associated with a stop sign 1134, a third landmark associated with a speed limit sign 1154, and a fourth landmark 1156 associated with a hazard sign 1138. Such landmarks may be used, for example, to assist an autonomous vehicle in determining its current position relative to any of the indicated target trajectories, and the vehicle may adjust its direction of travel to match the direction of the target trajectory at the determined position.

[0250] In some embodiments, the sparse map 800 may include a road signature profile. Such a road signature profile may be associated with any identifiable / measurable variation in at least one parameter associated with a road. For example, in some cases, such a profile may be associated with a variation in road surface information, such as a variation in the surface roughness of a particular road segment, a variation in the road width across a particular road segment, a variation in the distance between dotted lines painted along a particular road segment, or a variation in the road curvature along a particular road segment. Figure 11D shows an example of a road signature profile 1160. While the profile 1160 may represent any of the parameters listed above, in another example, the profile 1160 may represent a measurement of road surface roughness, such as one obtained by monitoring one or more sensors that provide an output indicating the amount of suspension slippage when a vehicle is traveling on a particular road segment.

[0251] Alternatively, or simultaneously, profile 1160 may represent variations in road width determined based on image data acquired via a camera mounted on a vehicle traveling along a particular road segment. Such a profile may be useful, for example, in determining a specific position of an autonomous vehicle relative to a particular target trajectory. That is, as the autonomous vehicle travels through a road segment, it may measure a profile associated with one or more parameters associated with the road segment. If the measured profile may correlate with / match a predetermined profile plotting parameter variations relative to a position along the road segment, the measured and predetermined profile may be used (for example, by superimposing the corresponding portions of the measured and predetermined profile) to determine the current position along the road segment, and therefore the current position relative to the target trajectory of the road segment.

[0252] In some embodiments, the sparse map 800 may include different trajectories based on different characteristics associated with the user of the autonomous vehicle, environmental conditions, and / or other parameters related to driving. For example, in some embodiments, different trajectories may be generated based on the preferences and / or profiles of different users. Sparse maps 800 including such different trajectories may be provided to different autonomous vehicles of different users. For example, some users may prioritize avoiding toll roads, while others may prioritize taking the shortest or fastest route, regardless of whether toll roads are present on the route. The disclosed system may generate different sparse maps using different trajectories based on such different user preferences or profiles. As another example, some users may prioritize driving in the high-speed lane, while others may prioritize always staying in the center lane.

[0253] Different trajectories may be generated and included in the sparse map 800 based on different environmental conditions, such as daytime and nighttime, snow, rain, fog, etc. An autonomous vehicle operating under different environmental conditions may be provided with a sparse map 800 generated based on such different environmental conditions. In some embodiments, a camera provided to the autonomous vehicle may detect environmental conditions and return such information to a server that generates and provides the sparse map. For example, the server may generate or update an already generated sparse map 800 to include trajectories that may be more appropriate or safer for autonomous driving under the detected environmental conditions. Updating the sparse map 800 based on environmental conditions may be performed dynamically as the autonomous vehicle travels along the road.

[0254] Other different parameters related to driving may be used as criteria for generating and providing different sparse maps for different autonomous vehicles. For example, turning may be more difficult when an autonomous vehicle is traveling at high speed. The sparse map 800 may include trajectories associated with a specific lane, rather than the road, so that the autonomous vehicle can stay within a specific lane when following a particular trajectory. If images captured by a camera mounted on the autonomous vehicle indicate that the vehicle has drifted outside the lane (e.g., crossed a lane marking), an action may be triggered within the vehicle to return the vehicle to the designated lane following a specific trajectory.

[0255] Crowdsourcing of sparse maps

[0256] In some embodiments, the disclosed systems and methods may generate sparse maps for the navigation of autonomous vehicles. For example, the disclosed systems and methods may use crowdsourced data to generate sparse maps that one or more autonomous vehicles can use to navigate along a system of roads. "Crowdsourcing" as used herein means that data is received from various vehicles (e.g., autonomous vehicles) traveling along a road segment at different times, and such data is used to generate and / or update a road model. The model may then be transmitted to vehicles traveling later along the road segment or to other vehicles to assist in the navigation of the autonomous vehicles. The road model may include a plurality of target trajectories representing preferred trajectories that autonomous vehicles should follow as they pass through the road segment. The target trajectories may be the same as reconstructed actual trajectories collected from vehicles traveling through the road segment, which can be transmitted from the vehicles to a server. In some embodiments, the target trajectories may differ from the actual trajectories previously taken by one or more vehicles as they pass through the road segment. The target trajectories may be generated based on the actual trajectories (e.g., through averaging or any other appropriate calculation).

[0257] Vehicle trajectory data that a vehicle may upload to a server may correspond to the vehicle's actual reconstructed trajectory, or to a recommended trajectory that may be based on or related to the vehicle's actual reconstructed trajectory but may differ from the actual reconstructed trajectory. For example, a vehicle may modify its actual reconstructed trajectory and submit (e.g., recommend) the modified actual trajectory to the server. A road model may use the recommended and modified trajectory as a target trajectory for the autonomous navigation of other vehicles.

[0258] In addition to trajectory information, other information that may be used when constructing the sparse data map 800 may include information related to potential landmark candidates. For example, through crowdsourcing of information, the disclosed systems and methods may identify potential landmarks in the environment and refine the location of the landmarks. Landmarks may be used by the navigation system of an autonomous vehicle to determine and / or adjust the vehicle's position along a target trajectory.

[0259] As a vehicle travels along a road, the reconstructed trajectory that the vehicle may generate may be obtained by any suitable method. In some embodiments, the reconstructed trajectory may be unfolded by combining segments of motion relative to the vehicle, for example, using ego-motion estimation (e.g., a camera, and therefore the three-dimensional translation and three-dimensional rotation of the vehicle body). Rotation and translation estimation may be determined based on the analysis of images captured by one or more image acquisition devices along with information from other sensors or devices, such as inertial sensors and velocity sensors. For example, the inertial sensor may include an accelerometer or other suitable sensor configured to measure changes during the translation and / or rotation of the vehicle body. The vehicle may include a velocity sensor to measure the vehicle's speed.

[0260] In some embodiments, the egomotion of a camera (and therefore the vehicle body) can be estimated based on optical flow analysis of the captured images. Optical flow analysis of a sequence of images identifies the motion of pixels from the sequence of images and, based on the identified motion, determines the motion of the vehicle. The egomotion can be integrated over time along the road segments to reconstruct the trajectory associated with the road segments followed by the vehicle.

[0261] Data collected at different times by multiple drives of multiple vehicles along a road segment (e.g., reconstructed trajectories) may be used to construct a road model (e.g., including target trajectories) contained in the sparse data map 800. Data collected by multiple vehicles in multiple drives along a road segment at different times may be averaged to improve the accuracy of the model. In some embodiments, data on the road geometry and / or landmarks may be received from multiple vehicles traveling through a common road segment at different times. Such data received from different vehicles may be combined to generate and / or update a road model.

[0262] The geometric shape of the trajectory reconstructed along the road segment (and the target trajectory as well) may be represented by a curve in three-dimensional space, or by a spline connecting three-dimensional polynomials. The curve of the reconstructed trajectory may be determined from the analysis of a video stream or multiple images captured by a camera mounted on the vehicle. In some embodiments, a position is identified in each frame or image located several meters ahead of the vehicle's current position. This position is where the vehicle is expected to travel within a predetermined time period. This operation may be repeated for each frame, and simultaneously, the vehicle may calculate the camera's egomotion (rotation and translation). In each frame or image, a short-range model of the desired path is generated by the vehicle in a reference frame mounted on the camera. The short-range models may be combined to obtain a three-dimensional model of the road in several coordinate frames, which may be arbitrary or predetermined coordinate frames. The three-dimensional model of the road may then be fitted by a spline containing, or connecting, one or more polynomials of appropriate order.

[0263] One or more detection modules may be used to complete a short-range road model in each frame. For example, a bottom-up lane detection module may be used. A bottom-up lane detection module may be useful when lane markings are drawn on the road. This module can find edges in the image and assemble them together to form lane markings. A second module may be used in conjunction with the bottom-up lane detection module. The second module is a terminal-to-terminal deep neural network that can be trained to predict an accurate short-range path from the input image. In both modules, the road model may be detected in the image's coordinate frame and transformed into a three-dimensional space that can be virtually mounted on the camera.

[0264] The method for modeling the reconstructed trajectory may involve the accumulation of errors that may contain noise components due to the integration of egomotion over long periods; however, such errors are not significant because the resulting model can provide sufficient accuracy for navigation on a local scale. Furthermore, it is possible to cancel out the integrated errors by using external information sources such as satellite imagery or geodetic surveys. For example, the disclosed system and method may use a GNSS receiver to cancel out the accumulated errors. However, GNSS positioning signals are not always available and accurate. The disclosed system and method may enable steering applications that do not heavily depend on the effectiveness and accuracy of GNSS positioning. In such systems, the use of GNSS signals may be limited. For example, in some embodiments, the disclosed system may use GNSS signals solely for the purpose of indexing a database.

[0265] In some embodiments, the distance range (e.g., local scale) relevant to the navigation steering application of an autonomous vehicle may be about 50 meters, 100 meters, 200 meters, 300 meters, etc. Primarily, such distances may be used because a geometric road model is used for two purposes: planning the previous trajectory and localizing the vehicle in the road model. In some embodiments, when the control algorithm maneuvers the vehicle according to a target point placed 1.3 seconds ahead (or any time such as 1.5 seconds, 1.7 seconds, 2 seconds, etc.), planning the task may use a model over a typical range 40 meters ahead (or any other suitable forward distance such as 20 meters, 30 meters, 50 meters, etc.). The localization task uses a road model over a typical range 60 meters behind the vehicle (or any other suitable distance, e.g., 50 meters, 100 meters, 150 meters, etc.) according to a method called "tail alignment" which is described in more detail in another column. The disclosed system and method can generate a geometric model with sufficient accuracy over a specific range, e.g., 100 meters, such that the planned trajectory does not deviate more than 30 cm from the center of the lane, for example.

[0266] As described above, the 3D road model can be constructed by detecting short-range segments and joining them together. The joining can be enabled by calculating a six-step ego-motion model using video and / or images captured by a camera, data from inertial sensors that reflect the movement of the vehicle, and the speed signal of the host vehicle. The accumulated error can be small enough for some local distance ranges, such as about 100 meters. All of this can be completed in a single drive over a specific road segment.

[0267] In some embodiments, multiple runs may be used to average the resulting models and further improve their accuracy. The same vehicle may travel the same route multiple times, or multiple vehicles may transmit the model data they collect to a central server. In either case, the matching procedure may be performed to identify overlapping models and enable averaging in order to generate a target trajectory. The constructed model (e.g., including the target trajectory) may be used for steering if it satisfies bundle criteria. Further runs may be used for further model improvement and to adapt to changes in infrastructure.

[0268] Sharing of driving experiences (e.g., sensing data) between multiple vehicles becomes possible when they are connected to a central server. Each vehicle client may store a partial copy of a universal road model that may be relevant to its current location. Bidirectional update procedures can be performed by the vehicle and the server. The small footprint concept described above enables the disclosed system and method to perform bidirectional updates using extremely narrow bandwidth.

[0269] Information relating to potential landmarks may be determined and transmitted to a central server. For example, the disclosed systems and methods may determine one or more physical characteristics of a potential landmark based on one or more images containing the landmark. Physical characteristics may include the physical size of the landmark (e.g., height, width), the distance from the vehicle to the landmark, the distance between landmarks relative to previous landmarks, the lateral position of the landmark (e.g., the position of the landmark relative to the driving lane), the GPS coordinates of the landmark, the type of landmark, and identifying information of text on the landmark. For example, a vehicle may analyze one or more images captured by a camera to detect potential landmarks such as speed limit signs.

[0270] A vehicle may determine the distance from the vehicle to a landmark based on the analysis of one or more images. In some embodiments, the distance may be determined based on the analysis of an image of the landmark using an appropriate image analysis method, such as a scaling method and / or an optical flow method. In some embodiments, the disclosed systems and methods may be configured to determine the type or classification of a possible landmark. If a vehicle determines that a particular possible landmark stored in a sparse map corresponds to a predetermined type or classification, it may be sufficient for the vehicle to communicate an indication of the landmark type or classification along with its location to the server. The server may store such indications. Other vehicles may then capture an image of the landmark, process the image (e.g., using a classifier), and compare the result of the image processing with an indication stored on the server regarding the type of landmark. There may be various types of landmarks, different types of landmarks may be associated with different types of data to be uploaded and stored on the server, vehicles with different processing may detect landmarks and communicate information about the landmarks to the server, and vehicles equipped with the system may receive landmark data from the server and use that landmark data to identify landmarks during autonomous navigation.

[0271] In some embodiments, multiple autonomous vehicles traveling on a road segment may communicate with a server. A vehicle (or client) may generate a curve representing its drive in an arbitrary coordinate frame (e.g., through EgoMotion integration). The vehicle may detect and locate landmarks within the same frame. The vehicle may upload the curve and landmarks to the server. The server may collect data from the vehicles across multiple drives and generate an integrated road model. For example, as discussed below with respect to Figure 19, the server may use the uploaded curves and landmarks to generate a sparse map with an integrated road model.

[0272] The server may supply models to clients (e.g., vehicles). For example, the server may supply sparse maps to one or more vehicles. The server may continuously or periodically update the models when it receives new data from vehicles. For example, the server may process new data and evaluate whether it contains information that should trigger an update or creation of new data on the server. The server may supply updated models or updates to vehicles to provide navigation for autonomous vehicles.

[0273] The server may use one or more criteria to determine whether new data received from a vehicle should trigger an update to the model or the creation of new data. For example, if new data indicates that a previously recognized landmark at a particular location no longer exists or has been replaced by another landmark, the server may determine that the new data should trigger an update to the model. As another example, if new data indicates that a road segment is closed, and this is confirmed by data received from another vehicle, the server may determine that the new data should trigger an update to the model.

[0274] The server may supply the updated model (or updated portion of the model) to one or more vehicles traveling on a road segment associated with the model update. The server may also supply the updated model to a vehicle that is scheduled to travel on a road segment associated with the model update, or to a vehicle whose planned travel includes a road segment. For example, while an autonomous vehicle is traveling along another road segment before reaching the road segment associated with the update, the server may supply the updated or updated model to the autonomous vehicle before the vehicle reaches the road segment.

[0275] In some embodiments, a remote server may collect trajectories and landmarks from multiple clients (e.g., vehicles traveling along a common road segment). Based on the trajectories collected from multiple vehicles, the server may use landmarks to match curves and create an average road model. The server may also calculate a road graph and the most probable path at each node, or the connections of road segments. For example, the remote server may align the trajectories and generate a crowdsourced sparse map from the collected trajectories.

[0276] The server may average landmark attributes received from multiple vehicles traveling along a common road segment, such as the distance between one landmark and another landmark (e.g., a previous one along the road segment), to determine the arc length parameter and support route localization and speed calibration for each client vehicle. The server may average the physical dimensions of a landmark measured by multiple vehicles traveling along a common road segment and recognizing the same landmark. The averaged physical dimensions may be used to support distance estimation, such as the distance from a vehicle to a landmark. The server may average the lateral position of a landmark (e.g., the position of the vehicle from the lane it is traveling in towards the landmark) measured by multiple vehicles traveling along a common road segment and recognizing the same landmark. The averaged lateral portion may be used to support lane designation. The server may average the GPS coordinates of a landmark measured by multiple vehicles traveling along the same road segment and recognizing the same landmark. The averaged GPS coordinates of the landmark may be used in the road model to support global localization or positioning of the landmark.

[0277] In some embodiments, the server may identify changes to the model, such as construction, detours, new signs, or removal of signs, based on data received from the vehicle. The server may continuously, periodically, or immediately update the model when it receives new data from the vehicle. The server may provide updates to the model or provide the vehicle with the updated model in order to provide autonomous navigation. For example, as will be discussed further below, the server may use crowdsourced data to remove "ghost" landmarks detected by the vehicle.

[0278] In some embodiments, the server may analyze driver interference during autonomous driving. The server may analyze data received from the vehicle at the time and location of the interference, and / or data received before the interference occurred. The server may identify specific portions of data that are caused by or closely related to the interference, such as data indicating a temporary lane closure or data indicating pedestrians on the road. The server may update the model based on the identified data. For example, the server may correct one or more trajectories stored in the model.

[0279] Figure 12 is a schematic diagram of a system that generates (and uses and navigates using) a sparse map using crowdsourcing. Figure 12 shows a road segment 1200 containing one or more lanes. Multiple vehicles 1205, 1210, 1215, 1220, and 1225 may travel on the road segment 1200 simultaneously or at different times (although shown to appear on the road segment 1200 at the same time in Figure 12). At least one of the vehicles 1205, 1210, 1215, 1220, and 1225 may be an autonomous vehicle. For simplicity in this example, we assume that all of the vehicles 1205, 1210, 1215, 1220, and 1225 are autonomous vehicles.

[0280] Each vehicle may be similar to a vehicle disclosed in other embodiments (e.g., vehicle 200) and may include components or devices included in or associated with a vehicle disclosed in other embodiments. Each vehicle may be equipped with an image acquisition device or camera (e.g., image acquisition device 122 or camera 122). Each vehicle may communicate with a remote server 1230 via one or more networks (e.g., via a cellular network and / or the Internet, etc.) through a wireless communication path 1235 as shown by a dotted line. Each vehicle may transmit data to and receive data from the server 1230. For example, the server 1230 may collect data from multiple vehicles traveling on a road segment 1200 at different times and may process the collected data to generate a road navigation model for an autonomous vehicle, or updates to the model. The server 1230 may transmit the road navigation model for an autonomous vehicle or updates to the model to the vehicles that transmitted data to the server 1230. The server 1230 may then transmit the autonomous vehicle's road navigation model or updates to the model to other vehicles traveling on the road segment 1200.

[0281] As vehicles 1205, 1210, 1215, 1220, and 1225 travel along road segment 1200, navigation information collected (e.g., detected, sensed, or measured) by vehicles 1205, 1210, 1215, 1220, and 1225 may be transmitted to server 1230. In some embodiments, the navigation information may be associated with a common road segment 1200. The navigation information may include trajectories associated with each of vehicles 1205, 1210, 1215, 1220, and 1225 as each vehicle travels across road segment 1200. In some embodiments, the trajectories may be reconstructed based on data detected by various sensors and devices provided to vehicle 1205. For example, the trajectory may be reconstructed based on at least one of accelerometer data, velocity data, landmark data, road geometry or profile data, vehicle position data, and ego-motion data. In some embodiments, the trajectory can be reconstructed based on data from inertial sensors, such as accelerometers, and the speed of the vehicle 1205 detected by a velocity sensor. Furthermore, in some embodiments, the trajectory can be determined (for example, by a processor mounted on each of the vehicles 1205, 1210, 1215, 1220, and 1225) based on the detected egomotion of a camera that may exhibit three-dimensional translation and / or three-dimensional rotation (or rotational motion). The egomotion of the camera (and therefore the vehicle body) can be determined from the analysis of one or more images captured by the camera.

[0282] In some embodiments, the trajectory of the vehicle 1205 may be determined by a processor mounted on the vehicle 1205 and transmitted to the server 1230. In other embodiments, the server 1230 may receive data detected by various sensors and devices provided to the vehicle 1205 and determine the trajectory based on the data received from the vehicle 1205.

[0283] In some embodiments, the navigation information transmitted from vehicles 1205, 1210, 1215, 1220, and 1225 to the server 1230 may include data relating to the road surface, the geometric shape of the road, or the road profile. The geometric shape of the road segment 1200 may include lane configuration and / or landmarks. The lane configuration may include the total number of lanes in the road segment 1200, the type of lane (e.g., one-way lane, two-way lane, driving lane, passing lane, etc.), markings on the lanes, lane width, etc. In some embodiments, the navigation information may include lane designation, for example, one lane out of several lanes in which the vehicle is traveling. For example, the lane designation may be associated with the number "3" indicating that the vehicle is traveling in the third lane from the left or right. As another example, the lane designation may be associated with the text value "center lane" indicating that the vehicle is traveling in the center lane.

[0284] Server 1230 may store navigation information on a non-temporary computer-readable medium, such as a hard drive, compact disk, tape, or memory. Based on navigation information received from multiple vehicles 1205, 1210, 1215, 1220, and 1225 (for example, through a processor included in Server 1230), Server 1230 may generate at least a portion of the autonomous vehicle's road navigation model for a common road segment 1200 and store the model as part of a sparse map. Based on crowdsourced data (e.g., navigation information) received from multiple vehicles (e.g., 1205, 1210, 1215, 1220, and 1225) traveling in the lanes of the road segment at different times, Server 1230 may determine the trajectories associated with each lane. Based on the multiple trajectories determined based on the crowdsourced navigation data, Server 1230 may generate the autonomous vehicle's road navigation model or a portion of the model (e.g., an updated portion). To update an existing autonomous vehicle road navigation model provided to the vehicle's navigation system, server 1230 may transmit the model or an updated portion of the model to one or more of the autonomous vehicles 1205, 1210, 1215, 1220, and 1225 traveling on road segment 1200, or to any other autonomous vehicles traveling on the road segment thereafter. The autonomous vehicle road navigation model may be used by the autonomous vehicles when autonomously navigating along the common road segment 1200.

[0285] As described above, the road navigation model of an autonomous vehicle may be contained in a sparse map (for example, sparse map 800 shown in Figure 8). Sparse map 800 may contain a sparse record of data related to the geometric shape of the road and / or landmarks along the road, and may provide sufficient information to guide the autonomous navigation of the autonomous vehicle without requiring excessive data storage. In some embodiments, the road navigation model of the autonomous vehicle may be stored separately from sparse map 800, and the model may use map data from sparse map 800 when it is run for navigation. In some embodiments, the road navigation model of the autonomous vehicle may use map data contained in sparse map 800 to determine a target trajectory along road segment 1200 in order to guide the autonomous navigation of autonomous vehicles 1205, 1210, 1215, 1220, and 1225, or other vehicles that subsequently travel along road segment 1200. For example, if the road navigation model for an autonomous vehicle is executed by a processor included in the vehicle 1205's navigation system, the model may cause the processor to compare a predetermined trajectory included in the sparse map 800 with a trajectory determined based on navigation information received from the vehicle 1205 in order to verify and / or correct the course the vehicle 1205 is currently traveling.

[0286] In a road navigation model for an autonomous vehicle, the geometric shape of road features or target trajectories can be encoded by curves in three-dimensional space. In one embodiment, the curve may be a three-dimensional spline containing one or more connected three-dimensional polynomials. Those skilled in the art will understand that, in order to fit the data, the spline may be a numerical function piecewise defined by a series of polynomials. Splines for fitting three-dimensional geometric shape data of roads may include linear splines (first order), quadratic splines (second order), three-dimensional splines (third order), or any other splines (other orders), or combinations thereof. The spline may contain one or more three-dimensional polynomials of different dimensions that connect (e.g., fit) the data points of the three-dimensional geometric shape data of roads. In some embodiments, a road navigation model for an autonomous vehicle may include three-dimensional splines corresponding to a common road segment (e.g., road segment 1200) or target trajectories along the lanes of road segment 1200.

[0287] As described above, the road navigation model of the autonomous vehicle included in the sparse map may include other information, such as identification information for at least one landmark along road segment 1200. The landmark may be visible within the field of view of a camera (e.g., camera 122) installed in each of the vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, camera 122 may capture an image of the landmark. A processor provided to vehicle 1205 (e.g., processors 180, 190, or processing unit 110) may process the image of the landmark to extract identification information about the landmark. The landmark identification information may be stored in the sparse map 800 rather than the actual image of the landmark. The landmark identification information may require considerably less storage space than the actual image. Other sensors or systems (e.g., a GPS system) may provide specific identification information for the landmark (e.g., the location of the landmark). A landmark may include at least one of the following: traffic signs, arrow marks, lane markings, dashed lane markings, traffic lights, stop lines, directional signs (e.g., highway exit signs with directional arrows, highway signs with arrows indicating different directions or locations), landmark beacons, or lampposts. A landmark beacon refers to a device (e.g., an RFID device) installed along a road segment that transmits or reflects signals to a receiver installed on a vehicle, and when a vehicle passes by the device, the beacon and the location of the device (e.g., determined from the device's GPS location) received by the vehicle may be used as a landmark to be included in the autonomous vehicle's road navigation model and / or sparse map 800.

[0288] The identification information for at least one landmark may include the location of at least one landmark. The location of a landmark may be determined based on position measurements performed using a sensor system associated with multiple vehicles 1205, 1210, 1215, 1220, and 1225 (e.g., a global positioning system, an inertia-based positioning system, a landmark beacon, etc.). In some embodiments, the location of a landmark may be determined by averaging position measurements detected, collected, or received by the sensor systems of different vehicles 1205, 1210, 1215, 1220, and 1225 over multiple drives. For example, vehicles 1205, 1210, 1215, 1220, and 1225 may transmit position measurement data to a server 1230, which may average the position measurements and use the averaged position measurements as the location of the landmark. The location of a landmark may be continuously improved by measurements received from the vehicles in subsequent drives.

[0289] Landmark identification information may include the size of the landmark. A processor provided to a vehicle (e.g., 1205) may estimate the physical size of the landmark based on image analysis. Server 1230 may receive multiple estimates of the physical size of the same landmark from different vehicles and different driving conditions. Server 1230 may average the different estimates to obtain the physical size of the landmark and store that landmark size in the road model. The physical size estimate may be used to further determine or estimate the distance from the vehicle to the landmark. The distance to the landmark may be estimated based on the vehicle's current speed and a scaling based on the location of the landmark appearing in the image relative to the camera's extended focus. For example, the distance to the landmark may be estimated by Z = V × dt × R / D, where V is the vehicle's speed, R is the distance in the image from the landmark to the extended focus at time t1, and D is the change in distance to the landmark in the image from t1 to t2. dt represents (t2-t1). For example, the distance to a landmark can be estimated by Z = V × dt × R / D, where V is the vehicle's speed, R is the distance in the image between the landmark and the extended focus, dt is the time interval, and D is the image displacement of the landmark along the epipolar line. Another equivalent formula, for example, Z = V × ω / Δω, can be used to estimate the distance to a landmark, where V is the vehicle's speed, ω is the image length (such as the width of an object), and Δω is the change in image length per unit time.

[0290] If the physical size of a landmark is known, the distance to the landmark may be determined based on the following formula, Z = f × W / ω, where f is the focal length, W is the size of the landmark (e.g., height or width), and ω is the number of pixels when the landmark leaves the image. From the above formula, the change in distance Z is ΔZ = f × W × Δω / ω 2It can be calculated using +f × ΔW / ω, where ΔW is attenuated to zero by averaging, and Δω is the number of pixels representing the bounding box accuracy of the image. The value for estimating the physical size of a landmark can be calculated on the server side by averaging multiple observations. The resulting error in distance estimation can be very small. There are two sources of error that can occur when using the above formula, i.e., ΔW and Δω. Their contribution to the distance error is given by ΔZ = f × W × Δω / ω 2 It is given by +f × ΔW / ω. However, ΔW is attenuated to zero by averaging, and therefore ΔZ is determined by Δω (e.g., the inaccuracy of the bounding box in the image).

[0291] For landmarks of unknown dimensions, the distance to the landmark can be estimated by tracking feature points in the landmark across a series of frames. For example, certain features appearing on a speed limit sign can be tracked across two or more image frames. Based on these tracked features, a distance distribution per feature point can be generated. Distance estimates can be extracted from this distance distribution. For example, the most frequent distance appearing in the distance distribution can be used as the distance estimate. Alternatively, the mean of the distance distribution can be used as the distance estimate.

[0292] Figure 13 shows an exemplary road navigation model for an autonomous vehicle represented by a plurality of three-dimensional splines 1301, 1302, and 1303. The curves 1301, 1302, and 1303 shown in Figure 13 are for illustrative purposes only. Each spline may contain one or more three-dimensional polynomials connecting a plurality of data points 1310. Each polynomial may be a linear polynomial, a quadratic polynomial, a cubic polynomial, or any suitable combination of polynomials having different dimensions. Each data point 1310 may be associated with navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, each data point 1310 may be associated with landmark-related data (e.g., landmark size, location, and identification information) and / or road signature profiles (e.g., road geometry, road roughness profile, road curvature profile, road width profile). In some embodiments, some data points 1310 may be associated with data related to landmarks, while others may be associated with data related to road signature profiles.

[0293] Figure 14 shows raw location data 1410 (e.g., GPS data) received from five separate drives. A drive may be separate from another drive if separate vehicles pass simultaneously, the same vehicle passes at different times, or separate vehicles pass at different times. To account for errors in the location data 1410 and different positions of vehicles within the same lane (e.g., one vehicle may be driving closer to the left side of the lane than another vehicle), the server 1230 may generate a map outline 1420 using one or more statistical techniques to determine whether the variation in the raw location data 1410 represents an actual difference or a statistical error. Each path in the outline 1420 may be traced back to the corresponding raw data 1410 that formed that path. For example, the path between A and B in the outline 1420 corresponds to raw data 1410 from drives 2, 3, 4, and 5, not from drive 1. Summary 1420 may not be detailed enough to be used for navigating vehicles (unlike the splines mentioned above, for example, as it combines driving from multiple lanes on the same road), but it can provide useful topological information and be used to define intersections.

[0294] Figure 15 shows an example in which further detail may be generated for a sparse map within a segment of the map overview (e.g., segment A to B within overview 1420). As shown in Figure 15, data (e.g., egomotion data and road marking data) may be indicated according to a location S (or S1 or S2) along the drive. Server 1230 may identify landmarks for the sparse map by identifying a unique match between landmarks 1501, 1503 and 1505 of drive 1510 and landmarks 1507 and 1509 of drive 1520. Such a matching algorithm may result in the identification of landmarks 1511, 1513 and 1515. However, those skilled in the art will recognize that other matching algorithms may be used. For example, stochastic optimization may be used instead of, or in combination with, unique matching. Server 1230 may align the drives longitudinally and align the matched landmarks. For example, server 1230 may select one operation (e.g., operation 1520) as the baseline operation, and then shift and / or flexibly extend other operations (e.g., operation 1510) for alignment.

[0295] Figure 16 shows an example of aligned landmark data used in a sparse map. In the example in Figure 16, landmark 1610 has a road sign. The example in Figure 16 further shows data from multiple drives 1601, 1603, 1605, 1607, 1609, 1611, and 1613. In the example in Figure 16, the data from drive 1613 consists of "ghost" landmarks, and none of drives 1601, 1603, 1605, 1607, 1609, and 1611 contain identification information for landmarks in the vicinity of the landmark identified in drive 1613, so server 1230 can identify it as such. Therefore, server 1230 may approve a potential landmark if the ratio of images with landmarks to images without landmarks exceeds a threshold, and / or may reject a potential landmark if the ratio of images without landmarks to images with landmarks exceeds a threshold.

[0296] Figure 17 shows a system 1700 for generating driving data, which may be used to crowdsource sparse maps. As shown in Figure 17, system 1700 may include a camera 1701 and a location device 1703 (e.g., a GPS locator). The camera 1701 and location device 1703 may be mounted on a vehicle (e.g., one of vehicles 1205, 1210, 1215, 1220, and 1225). Camera 1701 may provide multiple types of data, such as ego-motion data, traffic sign data, or road data. Camera data and location data may be segmented into driving segments 1705. For example, each driving segment 1705 may have camera data and location data from driving less than 1 km.

[0297] In some embodiments, the system 1700 can remove redundancy in the driving segment 1705. For example, if a landmark appears in multiple images from camera 1701, the system 1700 can remove redundant data so that the driving segment 1705 contains only a copy of either the location of the landmark or any associated metadata. As a further example, if lane markings appear in multiple images from camera 1701, the system 1700 can remove redundant data so that the driving segment 1705 contains only a copy of either the location of any metadata or any metadata associated with the lane markings.

[0298] System 1700 also includes a server (e.g., server 1230). Server 1230 can receive multiple driving segments 1705 from the vehicle and recombine them into a single driving segment 1707. Such a configuration may also allow the server to store data related to the entire driving segment while reducing the bandwidth requirements for moving data between the vehicle and the server.

[0299] Figure 18 shows the system 1700 of Figure 17, further configured for crowdsourcing sparse maps. As shown in Figure 17, the system 1700 includes a vehicle 1810 that captures driving data using, for example, a camera and a location device (e.g., a GPS locator) (providing, for example, ego-motion data, traffic sign data, or road data). As shown in Figure 17, the vehicle 1810 segments the collected data into multiple driving segments (shown in Figure 18 as "DS1 1", "DS2 1", and "DSN 1"). The server 1230 then receives the driving segments and reconstructs the driving (shown in Figure 18 as "Driving 1") from the received segments.

[0300] As further shown in Figure 18, system 1700 also receives data from additional vehicles. For example, vehicle 1820 also captures driving data using, for example, a camera (providing, for example, ego-motion data, traffic sign data, or road data) and a location device (for example, a GPS locator). Similar to vehicle 1810, vehicle 1820 segments the collected data into multiple driving segments (shown in Figure 18 as “DS1 2”, “DS2 2”, and “DSN 2”). Server 1230 then receives the driving segments and reconstructs the driving (shown in Figure 18 as “Driving 2”) from the received segments. Any number of additional vehicles may be used. For example, Figure 18 also includes “Vehicle N” which captures driving data (shown in Figure 18 as “DS1 N”, “DS2 N”, and “DSN N”), segments it into driving segments, and sends it to server 1230 to be reconstructed into driving (shown in Figure 18 as “Driving N”).

[0301] As shown in Figure 18, server 1230 can construct a sparse map (represented as "map") using reconstructed driving data (e.g., "driving 1", "driving 2", and "driving N") collected from multiple vehicles (e.g., "vehicle 1" (also referred to as vehicle 1810)), "vehicle 2" (also referred to as vehicle 1820), and "vehicle N").

[0302] Figure 19 is a flowchart illustrating an exemplary process 1900 for generating a sparse map for autonomous vehicle navigation along road segments. Process 1900 may be performed by one or more processing devices included in server 1230.

[0303] Process 1900 may include a step (step 1905) of receiving multiple images acquired as one or more vehicles pass through a road segment. Server 1230 may receive images from cameras contained within one or more of the vehicles 1205, 1210, 1215, 1220, and 1225. For example, camera 122 may capture one or more images of the environment surrounding vehicle 1205 as vehicle 1205 is traveling along road segment 1200. In some embodiments, server 1230 may receive deleted image data that had redundancy removed by the processor of vehicle 1205, as described above with respect to Figure 17.

[0304] The process 1900 may further include a step (step 1910) of identifying at least one line representation of a road surface feature extending along a road segment based on a plurality of images. Each line representation may represent a path along a road segment substantially corresponding to a road surface feature. For example, the server 1230 may analyze environmental images received from camera 122 to identify road edges or lane markings and determine the trajectory of travel along a road segment 1200 associated with the road edges or lane markings. In some embodiments, the trajectory (or line representation) may include a spline, a polynomial representation, or a curve. The server 1230 may determine the trajectory of travel of vehicle 1205 based on camera ego-motion (e.g., 3D translation and / or 3D rotation) received in step 1905.

[0305] Process 1900 may include a step (step 1910) of identifying multiple landmarks associated with a road segment based on multiple images. For example, server 1230 may analyze environmental images received from camera 122 to identify one or more landmarks, such as road signs along road segment 1200. Server 1230 may identify landmarks using analysis of multiple images acquired as one or more vehicles pass through the road segment. To enable crowdsourcing, the analysis may include rules for approving and rejecting landmarks that may be associated with a road segment. For example, the analysis may include approving a potential landmark if the ratio of images showing the landmark to images not showing the landmark exceeds a threshold, and / or rejecting a potential landmark if the ratio of images not showing the landmark to images showing the landmark exceeds a threshold.

[0306] Process 1900 may include other operations or steps performed by the server 1230. As will be discussed in more detail below, for example, the navigation information may include a target trajectory for a vehicle traveling along a road segment, and process 1900 may include the steps of the server 1230 clustering vehicle trajectories related to multiple vehicles traveling along the road segment, and determining the target trajectory based on the clustered vehicle trajectories. The step of clustering vehicle trajectories may include the server 1230 clustering multiple trajectories related to vehicles traveling along the road segment into multiple clusters based on at least one of the vehicle's absolute heading or the vehicle's lane designation. The step of generating the target trajectory may include the server 1230 averaging the clustered trajectories. As a further example, process 1900 may include the step of aligning the data received in step 1905. As described above, other processes or steps performed by the server 1230 may be included in process 1900.

[0307] The disclosed systems and methods may include other features. For example, the disclosed systems may use local coordinates rather than global coordinates. For autonomous driving, some systems may represent data in world coordinates. For example, longitude and latitude coordinates on the Earth's surface may be used. To use a map for steering, the host vehicle may determine its position and orientation relative to the map. To position the vehicle on the map, it would be natural to use an onboard GPS device to find rotational transformations between the vehicle reference frame and the world reference frame (e.g., north, east, and down). Once the vehicle reference frame is aligned using the map reference frame, a desired path may then be represented on the vehicle reference frame, and steering commands may be calculated or generated.

[0308] The disclosed systems and methods may enable autonomous vehicle navigation (e.g., steering control) using low-footprint models that can be collected by the autonomous vehicle itself without the use of expensive survey equipment. To support autonomous navigation (e.g., steering applications), the road model may include a sparse map with the road's geometry, lane configuration, and landmarks, which can be used to determine the location or position of a vehicle along a trajectory included in the model. As described above, the generation of the sparse map may be performed by a remote server that communicates with and receives data from a vehicle traveling on the road. The data may include detection data, a trajectory reconstructed based on the detection data, and / or a recommended trajectory that may represent a modified and reconstructed trajectory. As discussed below, the server may send the model back to the vehicle, or to other vehicles subsequently traveling on the road, for use in autonomous navigation.

[0309] Figure 20 shows a block diagram of server 1230. Server 1230 may include a communication unit 2005 which may include both hardware components (e.g., communication control circuit, switch, and antenna) and software components (e.g., communication protocol, computer code). For example, the communication unit 2005 may include at least one network interface. Server 1230 can communicate with vehicles 1205, 1210, 1215, 1220, and 1225 through the communication unit 2005. For example, server 1230 can receive navigation information transmitted from vehicles 1205, 1210, 1215, 1220, and 1225 through the communication unit 2005. Server 1230 can supply road navigation models for autonomous vehicles to one or more autonomous vehicles through the communication unit 2005.

[0310] Server 1230 may include at least one non-temporary storage medium 2010, such as a hard drive, compact disk, tape, etc. Storage device 1410 may be configured to store data, such as navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225, and / or a road navigation model of an autonomous vehicle generated by Server 1230 based on the navigation information. Storage device 2010 may be configured to store any other information, such as a sparse map (for example, the sparse map 800 described above with respect to Figure 8).

[0311] In addition to, or instead of, the storage device 2010, the server 1230 may include memory 2015. Memory 2015 may be similar to, or different from, memory 140 or 150. Memory 2015 may be non-temporary memory, such as flash memory or random access memory. Memory 2015 may be configured to store data, such as computer code or instructions executable by a processor (e.g., processor 2020), map data (e.g., data for sparse map 800), road navigation models for autonomous vehicles, and / or navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225.

[0312] Server 1230 may include at least one processing device 2020 configured to perform various functions by executing computer code or instructions stored in memory 2015. For example, processing device 2020 may analyze navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225 and generate a road navigation model for the autonomous vehicle based on the analysis. Processing device 2020 may control communication unit 1405 to supply the road navigation model for the autonomous vehicle to one or more autonomous vehicles (e.g., one or more of vehicles 1205, 1210, 1215, 1220, and 1225, or any vehicle that subsequently travels on road segment 1200). Processing device 2020 may be similar to or different from processors 180, 190, or processing unit 110.

[0313] Figure 21 shows a block diagram of memory 2015 that may store computer code or instructions for performing one or more operations to generate a road navigation model used in the navigation of an autonomous vehicle. As shown in Figure 21, memory 2015 may store one or more modules for performing operations to process the vehicle's navigation information. For example, memory 2015 may include a model generation module 2105 and a model supply module 2110. The processor 2020 may execute instructions stored in either module 2105 or 2110 contained in memory 2015.

[0314] The model generation module 2105, when executed by the processor 2020, can store instructions that can generate at least a portion of the autonomous vehicle's road navigation model for a common road segment (e.g., road segment 1200) based on navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225. For example, when generating the autonomous vehicle's road navigation model, the processor 2020 may cluster the vehicle trajectories along the common road segment 1200 into different clusters. Based on the clustered vehicle trajectories for each of the different clusters, the processor 2020 may determine a target trajectory along the common road segment 1200. Such operation may include a step in each cluster to find an intermediate or average trajectory of the clustered vehicle trajectories (e.g., by averaging the data representing the clustered vehicle trajectories). In some embodiments, the target trajectory may be associated with a single lane of the common road segment 1200.

[0315] A road model and / or sparse map may store trajectories associated with road segments. These trajectories may be referred to as target trajectories provided to autonomous vehicles for autonomous navigation. Target trajectories may be received from multiple vehicles, or may be generated based on actual trajectories or recommended trajectories (actual trajectories with some modifications) received from multiple vehicles. Target trajectories contained in the road model or sparse map may be continuously updated (e.g., averaged) using new trajectories received from other vehicles.

[0316] A vehicle traveling on a road segment may collect data using various sensors. This data may include landmarks, road signature profiles, vehicle motion (e.g., accelerometer data, velocity data), and vehicle position (e.g., GPS data), and may either reconstruct the actual trajectory itself or transmit the data to a server that reconstructs the actual trajectory relative to the vehicle. In some embodiments, the vehicle may transmit data related to the trajectory (e.g., curves in an arbitrary reference frame), landmark data, and lane designations along the travel route to the server 1230. In multiple drives, different vehicles traveling along the same road segment may have different trajectories. The server 1230 may, through clustering processing, identify the routes or trajectories associated with each lane from the trajectories received from the vehicles.

[0317] Figure 22 shows a process for clustering vehicle trajectories associated with vehicles 1205, 1210, 1215, 1220, and 1225 in order to determine a target trajectory for a common road segment (e.g., road segment 1200). The target trajectory or multiple target trajectories determined from the clustering process may be included in the autonomous vehicle's road navigation model or sparse map 800. In some embodiments, vehicles 1205, 1210, 1215, 1220, and 1225 traveling along road segment 1200 may transmit multiple trajectories 2200 to the server 1230. In some embodiments, the server 1230 may generate trajectories based on landmark, road geometry, and vehicle motion information received from vehicles 1205, 1210, 1215, 1220, and 1225. To generate a road navigation model for an autonomous vehicle, the server 1230 may cluster the vehicle's trajectory 1600 into multiple clusters 2205, 2210, 2215, 2220, 2225, and 2230, as shown in Figure 22.

[0318] Clustering can be performed using various criteria. In some embodiments, all driving within a cluster may be similar with respect to the absolute direction along road segment 1200. The absolute direction can be obtained from GPS signals received by vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, the absolute direction can be obtained using dead reckoning. As those skilled in the art will understand, dead reckoning can be used to determine the current position and, therefore, the direction of travel of vehicles 1205, 1210, 1215, 1220, and 1225, using previously determined positions, estimated speeds, etc. Trajectories clustered by absolute direction may be useful for identifying routes along the roadway.

[0319] In some embodiments, all driving within a cluster may be similar to lane designations along driving on road segment 1200 (e.g., in the same lane before and after a junction). Trajectories clustered by lane designations may be useful for identifying lanes along the roadway. In some embodiments, both criteria (e.g., absolute direction and lane designation) may be used for clustering.

[0320] In each cluster 2205, 2210, 2215, 2220, 2225, and 2230, trajectories may be averaged to obtain a target trajectory associated with a particular cluster. For example, trajectories from multiple drives associated with the same lane cluster may be averaged. The averaged trajectory may be a target trajectory associated with a particular lane. To average the clusters of trajectories, server 1230 may select a reference frame for any trajectory C0. For all other trajectories (C1, ...Cn), server 1230 may find a rigid transformation that maps Ci to C0, where i = 1, 2, ...n, and n is a positive integer corresponding to the total number of trajectories in the cluster. Server 1230 may calculate intermediate curves or trajectories within the C0 reference frame.

[0321] In some embodiments, landmarks may define arc lengths that match between different driving directions and may be used for aligning the trajectory with the lane. In some embodiments, lane markings before and after a junction may be used for aligning the trajectory with the lane.

[0322] To construct lanes from the trajectory, server 1230 may select a reference frame for any lane. Server 1230 may map partially overlapping lanes to the selected reference frame. Server 1230 may continue mapping until all lanes are within the same reference frame. Adjacent lanes may be aligned as if they were the same lane, and may be laterally shifted afterward.

[0323] Landmarks recognized along a road segment can be mapped to a common reference frame, first at the lane level and then at the junction level. For example, the same landmark may be recognized multiple times by multiple vehicles in multiple drives. Data on the same landmark received in different drives may differ slightly. Such data can be averaged and mapped to the same reference frame, e.g., the C0 reference frame. Alternatively, the variance of data on the same landmark received in multiple drives can be calculated.

[0324] In some embodiments, each lane of the road segment 120 may be associated with a target trajectory and a specific landmark. A target trajectory or a number of such target trajectories may be included in the road navigation model of an autonomous vehicle and may be later used by other autonomous vehicles traveling along the same road segment 1200. Landmarks identified by vehicles 1205, 1210, 1215, 1220, and 1225 while they are traveling along the road segment 1200 may be recorded in relation to the target trajectories. The target trajectory and landmark data may be continuously or periodically updated with new data received from other vehicles during subsequent driving.

[0325] To determine the position of an autonomous vehicle, the disclosed systems and methods may use an extended Kalman filter. The vehicle's position may be determined based on a prediction of its future position beyond its current position by integrating 3D positional data and / or 3D orientational data and egomotion. The vehicle's positioning may be modified or adjusted by image observation of landmarks. For example, if the vehicle detects a landmark in an image captured by a camera, the landmark may be compared to a known landmark stored in the road model or sparse map 800. The known landmark may have a known position (e.g., GPS data) along a target trajectory stored in the road model and / or sparse map 800. Based on the current speed and the image of the landmark, the distance from the vehicle to the landmark may be estimated. The vehicle's position along the target trajectory may be adjusted based on the distance to the landmark (stored in the road model or sparse map 800) and the known position of the landmark. It may be assumed that the location / positional data of the landmark stored in the road model and / or sparse map 800 (e.g., median values ​​from multiple drives) is accurate.

[0326] In some embodiments, the disclosed system may form a closed-loop subsystem in which the estimation of the vehicle's six degrees of freedom position (e.g., 3D position data, and further 3D directional data) may be used to navigate (e.g., steer the wheels) the autonomous vehicle to reach a desired location (e.g., stored 1.3 seconds ahead). Then, data measured from steering and actual navigation may be used to estimate the six degrees of freedom position.

[0327] In some embodiments, poles along the road, such as lampposts and power line poles or cable poles, may be used as landmarks to locate a vehicle. Other landmarks, such as traffic signs, traffic lights, road arrows, stop lines, and static features or signatures of objects along road segments, may also be used as landmarks to locate a vehicle. When poles are used to locate a vehicle, the bottom of the pole may be obstructed and may not be in the road plane, so x-observations of the pole (i.e., field of view from the vehicle) may be used rather than y-observations (i.e., distance to the pole).

[0328] Figure 23 shows a vehicle navigation system that may be used for autonomous navigation using a crowdsourced sparse map. For example, the vehicle is shown as vehicle 1205. The vehicle shown in Figure 23 may be any other vehicle disclosed herein, including, for example, vehicles 1210, 1215, 1220, and 1225, and vehicle 200 as shown in other embodiments. As shown in Figure 12, vehicle 1205 may communicate with server 1230. Vehicle 1205 may include an image acquisition device 122 (e.g., camera 122). Vehicle 1205 may include a navigation system 2300 configured to provide vehicle 1205 with navigation guidance for traveling on a road (e.g., road segment 1200). Vehicle 1205 may include other sensors, for example, a speed sensor 2320 and an accelerometer 2325. The speed sensor 2320 may be configured to detect the speed of vehicle 1205. The accelerometer 2325 may be configured to detect the acceleration or deceleration of the vehicle 1205. The vehicle 1205 shown in Figure 23 may be an autonomous vehicle, and the navigation system 2300 may be used to provide navigation guidance for autonomous driving. Alternatively, the vehicle 1205 may be a non-autonomous, human-controlled vehicle, and the navigation system 2300 may further be used to provide navigation guidance.

[0329] The navigation system 2300 may include a communication unit 2305 configured to communicate with the server 1230 via a communication path 1235. The navigation system 2300 may also include a GPS unit 2310 configured to receive and process GPS signals. The navigation system 2300 may further include at least one processor 2315 configured to process data, such as GPS signals, map data from a sparse map 800 (stored in a storage device mounted on the vehicle 1205 and / or received from the server 1230), the geometric shape of the road detected by the road profile sensor 2330, images captured by the camera 122, and / or a road navigation model of the autonomous vehicle received from the server 1230. The road profile sensor 2330 may include different types of devices for measuring different types of road profiles, such as road surface roughness, road width, road elevation, road curvature, etc. For example, the road profile sensor 2330 may include a device that measures the movement of the vehicle 2305's suspension to derive a road roughness profile. In some embodiments, the road profile sensor 2330 may include a radar sensor that measures the distance from the vehicle 1205 to the roadside (e.g., the roadside boundary), thereby measuring the width of the road. In some embodiments, the road profile sensor 2330 may include a device configured to measure the vertical elevation of the road. In some embodiments, the road profile sensor 2330 may include a device configured to measure the road curvature. For example, a camera (camera 122, or another camera) may be used to capture an image of the road showing the road curvature. The vehicle 1205 may use such an image to detect the road curvature.

[0330] At least one processor 2315 may be programmed to receive at least one environmental image associated with the vehicle 1205 from the camera 122. At least one processor 2315 may analyze at least one environmental image to determine navigation information associated with the vehicle 1205. The navigation information may include a trajectory related to the vehicle 1205's travel along the road segment 1200. At least one processor 2315 may determine the trajectory based on the operation of the camera 122 (and therefore the vehicle), e.g., three-dimensional translation and three-dimensional rotation. In some embodiments, at least one processor 2315 may determine the translation and rotation of the camera 122 based on the analysis of multiple images acquired by the camera 122. In some embodiments, the navigation information may include lane designation information (e.g., which lane the vehicle 1205 travels in along the road segment 1200). Navigation information transmitted from vehicle 1205 to server 1230 may be used by server 1230 to generate and / or update the road navigation model of the autonomous vehicle, and may be sent back from server 1230 to vehicle 1205 to provide autonomous navigation guidance to vehicle 1205.

[0331] At least one processor 2315 may be programmed to transmit navigation information from the vehicle 1205 to the server 1230. In some embodiments, the navigation information may be transmitted to the server 1230 along with road information. The road location information may include at least one of the following: GPS signals received by the GPS unit 2310, landmark information, road geometry, lane information, etc. At least one processor 2315 may receive a road navigation model or part of a model for the autonomous vehicle from the server 1230. The road navigation model for the autonomous vehicle received from the server 1230 may include at least one update based on the navigation information transmitted from the vehicle 1205 to the server 1230. The part of the model transmitted from the server 1230 to the vehicle 1205 may include the updated portion of the model. Based on the received road navigation model or updated portion of the model for the autonomous vehicle, at least one processor 2315 may enable the vehicle 1205 to perform at least one navigation operation (e.g., steering such as turning around, braking, accelerating, or overtaking another vehicle).

[0332] At least one processor 2315 may be configured to communicate with various sensors and B components included in the vehicle 1205, including a communication unit 1705, a GPS unit 2315, a camera 122, a speed sensor 2320, an accelerometer 2325, and a road profile sensor 2330. At least one processor 2315 may collect information or data from the various sensors and B components and transmit the information or data to the server 1230 via the communication unit 2305. Alternatively or further, various sensors or components of the vehicle 1205 may communicate with the server 1230 and transmit data or information collected by the sensors or components to the server 1230.

[0333] In some embodiments, vehicles 1205, 1210, 1215, 1220, and 1225 may communicate with each other and share navigation information, so that at least one of vehicles 1205, 1210, 1215, 1220, and 1225 may generate a road navigation model for the autonomous vehicle using crowdsourcing, for example, based on information shared by other vehicles. In some embodiments, vehicles 1205, 1210, 1215, 1220, and 1225 may share navigation information with each other, and each vehicle may update its own road navigation model for the autonomous vehicle provided to the vehicle. In some embodiments, at least one of vehicles 1205, 1210, 1215, 1220, and 1225 (e.g., vehicle 1205) may function as a hub vehicle. At least one processor 2315 of the hub vehicle (e.g., vehicle 1205) may perform some or all of the functions performed by the server 1230. For example, at least one processor 2315 of the hub vehicle may communicate with other vehicles and receive navigation information from them. At least one processor 2315 of the hub vehicle may generate a road navigation model for the autonomous vehicle, or update the model based on shared information received from other vehicles. At least one processor 2315 of the hub vehicle may transmit the road navigation model for the autonomous vehicle, or updates to the model, to other vehicles in order to provide autonomous navigation guidance.

[0334] Navigation parameter determination using mapping altitude

[0335] In many embodiments of autonomous or semi-autonomous navigation, determining the distance to objects in the host vehicle's environment can be useful. In some cases, this distance determination may be based on one or more outputs from a Lidar or Radar system. Alternatively, or further, the distance determination may be based on the analysis of images acquired from one or more image acquisition devices. However, in some cases, the analysis of acquired images can present certain difficulties. For example, distance measurement based on image analysis can be more difficult if objects in the host vehicle's environment have different altitudes relative to the host vehicle. For instance, in the analysis of acquired images, a target vehicle at a higher altitude than the host vehicle may be perceived as farther away than it actually is. To address this problem, distance determination based on image analysis may take into account the difference in altitude between the detected object and the host vehicle. This improves the accuracy of relative distance determination. Such altitude information can be determined in various ways. On the other hand, in some embodiments, altitude information may be stored for road segments (e.g., included in the sparse map described above). This mapping altitude information may be used to determine the distance between the host vehicle (particularly from a device that has acquired one or more images being analyzed) and a target object detected in the image. This technique, which considers altitude in distance measurement based on acquired images, is described in more detail below.

[0336] Figure 24 shows a block diagram of the memory, consistent with the disclosed embodiments. In some embodiments, memory 140 or 150 may include map information 2401, altitude information 2403, object identification module 2405, location module 2407, distance module 2409, and navigation module 2411. The disclosed embodiments are not limited to any particular configuration of memory 140 or 150. Furthermore, a processor (e.g., application processor 180, image processor 190, and / or processing unit 110) may execute instructions stored in any of the modules 2405, 2407, 2409, and 2411 contained in memory 140 or 150. Furthermore, although various modules (above and below) are described for performing functions related to the disclosed embodiments, these modules do not have to be logically separate and may instead be included in or integrated (in any combination) with an integrated navigation system configured to analyze acquired images and determine the next navigation state based on the characteristics of the host vehicle environment detected by the analysis of these images.

[0337] In one embodiment, the object identification module 2405 may store instructions that, when executed by the processor, enable the processor to perform image analysis on the image set and identify objects within the image set. In one embodiment, the object identification module 2405 may include a monocular image analysis module 402 and / or a stereo image analysis module 404. As described above, the monocular image analysis module 402 may include instructions for detecting a set of features in the image set, e.g., lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, hazardous materials, and any other features associated with the vehicle's environment. Also as described above, the stereo image analysis module 404 may include instructions for detecting a set of features in the first and second image sets, e.g., lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, hazardous materials, and any other features associated with the vehicle's environment. Based on these features, the object identification module 2405 may include instructions for identifying the type of object, such as a vehicle, pedestrians, objects in the roadway, manhole covers, traffic lights, traffic signs, etc. Furthermore, as part of the image analysis process, each identified object of a specific type may be assigned a bounding box associated with the edge boundaries of the object in the associated image.

[0338] In one embodiment, the positioning module 2407 may store instructions that, when executed by the processor, enable the processor to determine the position of the host vehicle. In one embodiment, the processor may determine the position of the host vehicle based on the output of at least one sensor of the host vehicle, such as a GPS device, a speed sensor, and / or an accelerometer. In one embodiment, the processor may determine the position of the host vehicle based on positioning on a sparse map road model (as described above). A sparse map not inconsistent with this disclosure may include data representing one or more road features. Such road features may include recognized landmarks, road signature profiles, and any other road-related features useful for navigating a vehicle. Using recognized landmarks identified in one or more acquired images, along with one or more target trajectories included in the sparse map, the processor may determine the position of the host vehicle relative to the target trajectories in the map. Such position information may inform the host vehicle of specific features associated with the host vehicle's environment relative to the determined position of the host vehicle. For example, if the location of a depression along a road segment is stored in a sparse map, then by determining the host vehicle's position relative to a specific target trajectory, it may be possible to determine that there is a depression 114 meters ahead of the host vehicle along the current target trajectory. Similarly, if road elevation information is stored in the sparse map (e.g., every 10 cm, 1 m, 2 m, 5 m, 10 m, 50 m, 100 m, etc.), the processor may determine that the elevation of the road segment 128 m ahead of the host vehicle's current specific position relative to a specific target trajectory is 27 m higher than the elevation of the road segment at the host vehicle's current specific position. After determining the host vehicle's position, as described above, the processor may estimate its position relative to the target trajectory (e.g., between recognized landmark identifications) based on sensor outputs indicating the host vehicle's ego-motion between recognized landmarks.

[0339] In one embodiment, map information 2401 (including, for example, the sparse map described above) may store features associated with a particular road segment. Features may include detected lane markings, road signs, highway exit ramps, traffic lights, road infrastructure, road width information, guardrails, trees, buildings, lampposts, and any other features related to the vehicle environment. In one embodiment, map information 2401 may further include elevation information 2403. Elevation information 2403 may include the height of the road segment at a particular location (e.g., 10 meters, 20 meters, 30 meters, etc.) or the gradient (e.g., 10 degrees or 11 degrees, etc.). In one embodiment, elevation information 2403 may include the height relative to an origin (e.g., sea level, or a specific origin associated with a particular road segment). In one embodiment, elevation information 2403 may be stored as a mathematical formula (e.g., formula (2)) each corresponding to a road segment. In one embodiment, altitude information 2403 may be stored in a table where each altitude information entry corresponds to a specific location along a road segment. Such altitude information may be stored in a sparse map at any appropriate interval (e.g., every 10 cm, 1 m, 2 m, 5 m, 10 m, 50 m, 100 m, etc.). In one embodiment, referring again to Figure 13, a plurality of three-dimensional splines 1301, 1302, and 1303 may be stored along with altitude information for lanes at a plurality of data points 1310. The processor may retrieve altitude information from the sparse map in any appropriate way suitable for a particular altitude storage protocol (e.g., reading altitude information from a table, determining altitude from a stored profile or function / formula, determining altitude from the z-axis projection of a three-dimensional spline used to indicate a target trajectory, etc.).

[0340] In one embodiment, the distance module 2409 may store instructions configured, when executed by the processor, to determine the distance from the host vehicle to an identified object. The distance may be determined based on measurements from image analysis, such as scaling and / or optical flow. In some cases, distance determination based on image analysis may be redundant to or used for verification of one or more distance determinations based on the output of a LIDAR or RADAR system. The determined distance from the host vehicle to an identified object in one or more images may depend on the altitude of the identified object relative to the host vehicle. For example, if the altitude of the target vehicle is higher than that of the host vehicle, the determined distance may be longer than the actual distance if altitude is not taken into account. Accurate distance values ​​to the detected object should be available to determine appropriate navigation actions for the host vehicle.

[0341] Without being inconsistent with the disclosed embodiments, the processor may determine the distance from the host vehicle to an identified object based on altitude information 2403. Figures 25A and 25B are schematic diagrams illustrating a situation where the target vehicle 2503 is located on a road segment at a higher altitude than the host vehicle 2501. For example, it can be assumed that the host vehicle 2501 is on a ground plane and the target vehicle 2503 is located on a plane at a higher elevation than the ground plane. The height Y, which is the height from the ground plane to the plane on which the target vehicle 2503 is located, can be expressed using the following equation.

number

[0342] In the formula, y represents the vertical displacement of the bottom of the target vehicle 2503 in image space, f represents the focal length of the image acquisition device, and Z represents the distance from the host vehicle to the target vehicle.

[0343] Furthermore, based on the road model, the height Y can be determined from the following equation.

number

[0344] In the equation, a represents a predetermined constant, and H represents the height of the image acquisition device relative to the ground plane. In one embodiment, the constant a can be stored in map information, a sparse map, and / or a three-dimensional spline. In some cases, the constant a may be a local approximation of the spline. In one embodiment, equation (2) may include a more complex function (e.g., Y = function (Z)).

[0345] In some cases, the processor may use equations (1) and (2) to determine the distance Z, which is the distance from the host vehicle 2501 to the target vehicle 2503. For example, the proximity region may be approximated as a plane. The vanishing of the plane (horizon) may be determined from the vanishing point of the proximity portion of the road. Alternatively, the local road plane may be determined by any appropriate image analysis technique. The bottom of the vehicle in the image relative to the "horizon" can define a spatial line relative to the bottom of the vehicle, i.e., Y = -yZ / f. This yields the following equation.

number

[0346] In one embodiment, the navigation module 2411 may be processor-executable and store software instructions that determine a desired navigation action based on a determined distance. In one embodiment, the navigation module 2411 includes a navigation response module 408. The determined distance from the host vehicle to an identified object (e.g., a target vehicle) may be important, for example, for generating an appropriate navigation response for the autonomous vehicle. Based on the determined distance, a navigation action may be determined to achieve one or more navigation goals for the autonomous vehicle (e.g., drive from point A to point B, avoid hitting pedestrians, maintain a distance of at least 1 meter from pedestrians, maintain a distance of at least 10 meters from pedestrians if the speed exceeds 30 km / h (etc.), maintain at least a safe distance from a preceding vehicle, apply the brakes to avoid approaching within a safe distance, and many other navigation goals). The processor may achieve these goals by relying on accurate distance determination between at least the host vehicle and the detected object. For example, if the host vehicle approaches a detected pedestrian, the braking system may be triggered to stop the host vehicle at a safe distance from the pedestrian. In one embodiment, the processor may compare the determined distance to a predetermined threshold. If the determined distance exceeds the predetermined threshold (for example, a safe distance to a preceding vehicle that can be the sum of the distance the host vehicle can travel at maximum acceleration relative to the reaction time associated with the host vehicle, the distance range at which the host vehicle can stop at its maximum brake level, the distance range at which the target vehicle can stop at its maximum brake level, and optionally, a predetermined minimum proximity buffer for maintaining distance from the target vehicle after stopping), the processor may allow the host vehicle to maintain its current speed, course, acceleration, etc. However, if the determined distance is equal to or lower than the predetermined threshold, the processor may cause the host vehicle to brake, change direction, etc. In some embodiments, the navigation response may be determined based on the determined distance value and a combination of relative speed and acceleration.

[0347] Figure 26 is a flowchart of host vehicle navigation, consistent with the disclosed embodiments. In step 2601, the processor may receive at least one image representing the host vehicle environment from an image acquisition device. This image may be raw or processed data sent from image acquisition devices 122, 124, and / or 126 via a network or communication path. The data may include any suitable image format.

[0348] In step 2603, the processor may analyze at least one image to identify objects in the host vehicle's environment. Image 2500 shows the host vehicle's environment, including numerous different objects. The processor may analyze image 2500 to identify objects in the image. The processor may access the object identification module 2405 to utilize any suitable image analysis technique. Examples of such techniques include object recognition, image segmentation, feature extraction, optical character recognition (OCR), object-based image analysis, shape domain techniques, edge detection methods, and pixel-based detection. In addition, the processor may further use classification algorithms to distinguish different objects in the image. In one embodiment, the processor may use appropriately trained machine learning algorithms and models to perform object identification. Algorithms may include linear regression, logistic regression, linear discriminant analysis, classification trees and regression trees, naive Bayes, k-nearest neighbors, learning vector quantization, support vector machines, bagging and random forests, and / or boosting and Adaboost. In some embodiments, the processor may identify objects in an image when the object identification module 2405 is running, based at least on the visual characteristics of the objects (e.g., size, shape, texture, characters, color, etc.).

[0349] In step 2605, the processor may determine the position of the host vehicle. In one embodiment, the processor may determine the position based on image analysis. Furthermore, as described above, the processor may determine the position of the host vehicle as a specific location of the host vehicle along a three-dimensional spline representing the target trajectory of the host vehicle along a road segment. Furthermore, without inconsistency with the disclosed embodiments, the prediction of the host vehicle's position along a predetermined three-dimensional spline may be based on the observed location of at least one recognized landmark. Without inconsistency with the disclosed embodiments, the processor may predict the position of the host vehicle based on localization on a sparse map road model. In one embodiment, as described above, the position of the host vehicle may be determined based on the output of at least one sensor of the host vehicle, such as a GPS device, a speed sensor, and / or an accelerometer. For example, location information from a GPS receiver may contribute to determining the position of the host vehicle.

[0350] In step 2607, the processor may receive map information associated with the determined location of the host vehicle. This map information includes altitude information associated with the host vehicle's environment. Based on the determined location (e.g., GPS output or previous location determination on a sparse map), the processor may obtain map information for a specific location or a specific road segment. The obtained map information may include altitude information for the road segment on which the host vehicle is traveling. The altitude information may be stored in the map, for example, in any of the methods described above.

[0351] In step 2609, the processor may determine the distance from the host vehicle to the object based on at least altitude information. According to the disclosed embodiments, the processor may use the above formula to determine the distance to an object identified in one or more images. For example, the relative distance to a target object may be calculated using formula (2) based on altitude information from a stored map.

[0352] In step 2611, the processor may determine a navigation action for the host vehicle based on the determined distance. In one embodiment, the processor may access the navigation module 2411 to determine a navigation action. The navigation action may include any of the above-described navigation actions, which may depend on the determined distance to the identified target object.

[0353] In some cases, as described above, the determined navigation action may include one or more navigation actions deployed in accordance with the objective of the host vehicle maintaining at least a safe distance from other vehicles. For example, if the determined distance, taking into account the altitude difference between the host vehicle and the identified target vehicle, is less than or equal to the safe distance, or if the host vehicle is close to the safe distance buffer zone, the navigation processor may determine a navigation action for the host vehicle aimed at achieving or maintaining at least a safe distance from the identified target vehicle. For example, in some cases, the safe distance may be the sum of the host vehicle's current stopping distance based on its current speed and current maximum braking performance, the acceleration distance corresponding to the distance the host vehicle travels at its current maximum acceleration performance during the reaction time associated with the host vehicle, and the stopping distance until the target vehicle reduces its current speed to zero due to the target vehicle's maximum braking performance. The navigation action may be determined based on one or more of the formulas described above in relation to the Responsibility-Based Safety Theory (RSS) model.

[0354] In some embodiments, the processor may perform a determined navigation operation if, based on the host vehicle's maximum braking performance, current speed, maximum acceleration performance, and / or reaction time associated with the host vehicle, the host vehicle can stop within a stopping distance less than the determined next-state distance (i.e., the distance between the host vehicle and the target vehicle as a result of performing the determined navigation operation). In some cases, the stopping distance of the target vehicle, based on its determined speed and maximum braking performance, may also be considered when performing the determined navigation operation. Assuming that there is no collision between the preceding visible target vehicle and the host vehicle, or that a collision for which the host vehicle is responsible occurs, and that the preceding visible target vehicle could suddenly come to a complete stop at any time, the host vehicle processor may perform the navigation operation as planned if it determines that there is a sufficient distance to stop at the next-state distance. On the other hand, if the distance is insufficient to stop the host vehicle without a collision or without the host vehicle being responsible for a collision, the determined navigation operation does not need to be performed.

[0355] Navigation parameter determination using mapped lane widths

[0356] Similar to the embodiments described above in which elevation information can be stored in a sparse map for a specific road segment, lane widths measured for a road segment can also be stored in a sparse map. For example, lane width determination may be performed along the road segment as a vehicle passes through it. These lane width measurements can be refined and then integrated into the sparse map described above by crowdsourcing. For example, the lane width information can be stored as a table along the longitudinal direction of the road segment, as a function or expression defining the width along the longitudinal direction with respect to one or more 3D splines indicating a target trajectory for lanes traveling along the road segment, or in any other appropriate form. As will be described in more detail later, the mapped lane width information can be useful for determining the distance between a host vehicle and one or more target objects identified on the road segment. The determined distance can be useful for verifying measurements from a LIDAR system, a RADAR system, or for determining one or more navigation actions for the host vehicle.

[0357] As described above, in one embodiment, the processor may receive one or more acquired images representing the surrounding environment of the host vehicle. In one embodiment, as described above, the object identification module 2405 may store instructions, when executed by the processor, that enable the processor to perform image analysis of the image set and identify objects within the image set. The processor may identify objects in multiple images, multiple landmarks associated with road segments (e.g., road signs). Objects may be identified by any of the techniques disclosed above. For example, object recognition, image segmentation, feature extraction, optical character recognition (OCR), object-based image analysis, shape region techniques, edge detection methods, pixel-based detection, etc., can all be used individually or in combination.

[0358] In one embodiment, as described above, the processor may determine the location of the host vehicle. As described above, the location may be determined by GPS localization, localization of recognized landmarks along a target trajectory on a sparse map, a combination of these, and various other techniques. For example, referring to Figure 27A, the specific location of the host vehicle 2701 along a spline representing a target trajectory associated with a lane on a road segment in which the host vehicle 2701 is traveling may be determined using landmark 2705, using the techniques described above.

[0359] In one embodiment, as described above, the distance module 2409 may store instructions that, when executed by the processor, enable the processor to determine the distance from the host vehicle to an identified object. For example, referring to Figure 27A, a host vehicle 2701, a target vehicle 2703, a landmark 2705, and two splines S1 and S2 are shown in three-dimensional space with (X,Y,Z) coordinates. The splines S1 and S2 may represent lane boundaries associated with the road segment on which the host vehicle 2701 is traveling. The distance determination between the host vehicle 2701 and the target vehicle 2703 may be based on the output of a LIDAR or RADAR system. The distance between the host vehicle 2701 and the target vehicle 2703 may further be determined by image analysis, for example, utilizing the scaling of objects between a series of images. The distance determination may take altitude differences into account, as described in the above section. Distance determination may also be based on the tracking trajectory of the target vehicle 2703, and the observed tracking trajectory may be compared with mapping information (such as splines representing the driving lane) to determine where the target vehicle is located along the target trajectory. The distance between the host vehicle 2701 and the target vehicle 2703 may be determined as the shortest straight-line distance between the host vehicle and the target vehicle. In other cases, the distance between the host vehicle 2701 and the target vehicle 2703 may be determined as the distance of the curved three-dimensional path between the host vehicle and the target vehicle (for example, along the splines between lane edge splines S1 and S2, representing the target trajectory that the host vehicle travels).

[0360] The processor may further determine the distance from the host vehicle to an identified object based on the mapped lane width information. For example, the map information in the sparse map described above may include lane width information associated with the road segment on which the host vehicle is traveling. As described above, the sparse map may store the lane width information in any suitable format (e.g., a relational table, periodic values ​​along the longitudinal direction of the road segment, or a three-dimensional spline capable of determining the lane width, such as between splines S1 and S2 shown in Figures 27A and 27B). For example, the lane width between splines S1 and S2 over a length distance associated with the target vehicle 2703 may be represented as the difference between the x-coordinate projection at point y2 associated with spline S2 and the x-coordinate projection at point y1 associated with spline S1.

[0361] Referring to Figure 28A, a host vehicle 2801, a target vehicle 2803, a landmark 2805, and two splines S1 and S2 are shown. Splines S1 and S2 may represent lane boundaries associated with the road segments on which the host and target vehicles are traveling. As shown in the figure, there is a width W between splines S1 and S2. In some embodiments, the width W may be constant or approximately constant between the two splines representing the detected lane boundaries. Figure 28B shows splines S1 and S2 projected onto a (x,y) two-dimensional coordinate system (e.g., in image space). Based on known lane width information (e.g., W = 3.8 meters), the distance Z from the host vehicle 2801 to the target vehicle 2803 can be determined by the following equation.

number

[0362] In the formula, f represents the focal length of the image acquisition device, and w represents the width of the lane in image space. For example, w could be 125 pixels and f could be 1400 pixels. If the lane width W in physical space is known to be 3.8 meters (for example, based on a stored sparse map), then Z can be calculated as 42.56 meters.

[0363] Figure 29A shows another image containing splines s1, s2, and s3 in a (x,y) 2D coordinate system. Figure 29B shows images of splines S1, S2, and S3 in 3D coordinates. In one embodiment, the server may receive the acquired image 2900 and use that image to construct image 2910 by mapping the (x,y) 2D coordinates to a (X,Y,Z) 3D coordinate system. For example, to construct image 2910, the server may detect a target vehicle 2903 and determine the lowest (x,y) value of the target vehicle 2903. In some embodiments, the server may apply a mapping function to calculate the (X,Y,Z) value of a point (x,y). Thus, a 3D spline can be constructed based on a 2D image.

[0364] In one embodiment, the server can generate a road surface model using all constructed splines. For example, using image 2910, the server can generate a road surface model which can be used as a segment of a map. For example, to generate a road surface model, the server can select points on splines S1, S2, and S3 at predetermined intervals (e.g., 1-meter intervals). Furthermore, as shown in Figure 29B, the server can apply Delaunay triangulation. Figure 30 shows an exemplary spline projected onto two-dimensional coordinates. Figure 31 shows the result of Delaunay triangulation. As shown, points outside the road edge 3101 are filtered out. Furthermore, region 3103 is added as a fixed-width margin outside the road edge. These (X,Y,Z) three-dimensional coordinates can be constructed by the server from the (x,y) two-dimensional coordinates as described above. From these three-dimensional values, the server can obtain information about the road 3101 and the area near the road (e.g., region 3103).

[0365] In some embodiments, the server may use one or more 3D splines from a map to map coordinates (e.g., position and rotation information) to map (x,y) 2D coordinates to a (X,Y,Z) 3D coordinate system. For example, the server may project 3D splines representing lane boundaries into an image. For all 3D points along the splines (e.g., spaced apart), the server may use known information about one or more cameras that captured the image (e.g., focal length, principal point, lens distortion, etc.) to calculate the 2D image projection. Instead of determining RGB values, the server may determine and store the (X,Y,Z) values ​​of the projection points. Thus, the server may determine (X,Y,Z) values ​​for all (x,y) 2D coordinates along the projected splines in the image. When the server detects a vehicle and its contact point with the road (e.g., detected in the image data), the server may examine points y1 and y2 on nearby splines s1 and s2 and determine the distance to the vehicle as a (weighted) average of y1 and y2. This method may be effective when the vehicle is equipped with multiple surround cameras.

[0366] Figure 32 shows the associated triangles projected onto the image based on the camera's position. When the server applies the Delaunay triangulation function, if a point in the image falls within a triangle, it assigns a (X,Y,Z) 3D value to that point based on that triangle.

[0367] In one embodiment, the host vehicle's processor may receive a surface model. As described above, the processor may use the techniques described above to determine the position of the host vehicle and its position in a three-dimensional coordinate system. For example, such position determination may be based on positioning relative to a identified landmark and / or positioning based on the determined ego-motion of the host vehicle relative to an identified object (e.g., a target vehicle) between the landmarks. The processor may use a road surface model to determine the distance from the host vehicle to the identified object. For example, when a target vehicle 2903 is detected, the point of contact (x,y) with the road becomes a spatial line. The 3D intersection point of this line with the road surface gives the distance Z from the host vehicle to the target vehicle.

[0368] As described above, the processor may determine navigation actions based on at least the determined distance. For example, if the determined distance from the host vehicle to the target vehicle is less than the current stopping distance, the processor may decide to apply the brakes.

[0369] Figure 33 is a flowchart illustrating an exemplary process for navigating a host vehicle, consistent with the disclosed embodiments. In step 3301, the processor may receive at least one image representing the environment of the host vehicle from an image acquisition device. This image may be raw or processed data sent from image acquisition devices 122, 124, and / or 126 via a network or communication path. The data may include data described in some image format.

[0370] In step 3303, the processor may analyze at least one image to identify objects in the host vehicle environment. Image 2800 shows the host vehicle environment, including numerous different objects. The processor may analyze image 2800 to identify objects in the image. As described above, the processor may access the object identification module 2705 and utilize any appropriate image analysis technique. In addition, the processor may further use classification algorithms to distinguish different objects in the image. In one embodiment, the processor may use appropriately trained machine learning algorithms and models to perform object identification.

[0371] In step 3305, the processor may determine the position of the host vehicle. In one embodiment, as described above, the processor may determine the position of the host vehicle based on a prediction of the host vehicle's position along a predetermined three-dimensional spline representing the target trajectory of the host vehicle along a road segment. For example, referring to Figure 27A, the processor may determine the position of the host vehicle based on an identified landmark 2705. Furthermore, without inconsistency with the disclosed embodiments, the prediction of the host vehicle's position along the predetermined three-dimensional spline may be based on the observed position of at least one recognized landmark. In one embodiment, as described above, the position of the host vehicle may be determined based on the output of at least one sensor of the host vehicle, such as a GPS device, a speed sensor, and / or an accelerometer.

[0372] In step 3307, the processor may access map information associated with the determined location of the host vehicle. This map information includes lane width information associated with roads within the host vehicle's environment. In one embodiment, the processor may receive a sparse map associated with a particular segment on which the host vehicle is traveling. Next, as described above, the processor may determine the host vehicle's position relative to a trajectory in the map. Once the trajectory is determined, the processor may determine the host vehicle's future path based on the determined specific location of the host vehicle along a three-dimensional spline representing its target trajectory. Furthermore, the processor may receive lane width information at all points (or at least equally spaced points) along the road segment ahead of the host vehicle. The lane width information may be stored in the sparse map in any suitable format. Examples of formats are as described above. In some embodiments, the lane width information may include left and right values ​​relative to the target trajectory spline at predetermined longitudinal distance values ​​along the spline, and / or any other arbitrary storage format.

[0373] In step 3309, the processor may determine the distance from the host vehicle to the object based on at least lane width information. For example, as described above, the processor may take the average of the y values ​​when the spline is projected onto a two-dimensional coordinate system (e.g., Z = (y1 + y2) / 2). In another example, the processor may use equation (4) to determine the distance from the host vehicle to the object. Alternatively, the distance may be determined based on the output of at least one sensor of the host vehicle, such as a LIDAR system, RADAR system, GPS device, speed sensor, and / or accelerometer. In one embodiment, an image analysis method that determines the distance from the host vehicle to the object based on at least lane width information may be used to verify the distance determined based on the output of another sensor of the host vehicle.

[0374] In step 3311, the processor may determine a navigation action for the host vehicle based on the determined distance. For example, as described above, the navigation action may be determined to maintain or achieve a safe distance (e.g., with respect to the RSS model as described above) from one or more detected vehicles.

[0375] Lane position determination for target vehicles that are partially invisible

[0376] In some cases, the entire target vehicle may be represented in the image acquired by the image acquisition device associated with the host vehicle. In such cases, the distance to the target vehicle may be determined based on the observed line or position where the target vehicle is in contact with the road surface in the image. However, in other cases, the entirety of one or more target vehicles may not be represented in the acquired image. For example, a portion of a target vehicle may be obscured from the acquired image, such that the intersection of one or more target vehicles with the road surface is not observable. In such cases, another technique may be used to estimate the positional information of one or more detected target vehicles and to determine the distance between the host vehicle and the detected target vehicles. Alternatively or further, the estimated positional information regarding the detected target vehicles can be used in combination with optionally stored map information (e.g., from the sparse map described above) to determine the lane in which the detected target vehicles are traveling, the expected direction of travel of the target vehicles, etc.

[0377] As described above, in one embodiment, a processor for image processing may receive one or more acquired images representing the surrounding environment of a host vehicle. In one embodiment, as described above, the processor may perform image analysis on the set of images to identify objects within the set of images. The processor may identify objects in multiple images, multiple landmarks associated with road segments (e.g., road signs). Objects may be identified by any of the techniques disclosed above, such as object recognition, image segmentation, feature extraction, optical character recognition (OCR), object-based image analysis, shape region techniques, edge detection methods, and pixel-based detection.

[0378] In one embodiment, based on analysis, the processor may identify a first object in the environment (e.g., a target vehicle). The first object and the road on which the first object is located are partially obscured by a second object (e.g., an obstruction, a road sign, etc.). Figure 34A shows an exemplary situation in an image captured from the host vehicle 3401 where the target vehicle may be at least partially obscured. Figure 34A shows the position of the host vehicle 3401 relative to the target vehicle 3403 and other objects such as an obstruction 3405.

[0379] Figure 34B shows an exemplary image that may be acquired by a camera associated with the host vehicle 3401. In the image shown in Figure 34B, the target vehicle 3403 can be detected in the image. However, the image of the target vehicle 3403 is partially obscured by the representation of an obstruction 3405. For example, the bottom of the target vehicle 3403 is obscured by the obstruction 3405. Nevertheless, the relevant lane in which the target vehicle is traveling (or the path in which the target vehicle is traveling and is expected to travel) can be determined, for example, by tracking the movement of the target vehicle 3403 over time (represented, for example, by two or more acquired images) and comparing it with mapping information (for example, the sparse map described above) to estimate the lane or path in which the target vehicle 3403 is traveling.

[0380] More specifically, based on two or more acquired images, the processor may determine the lane position and / or the position of the target vehicle relative to the stored map information. For example, based on a sparse map that stores trajectories for drivable lanes along a road segment (which may include multiple intersecting road sections), the processor may determine that the target vehicle 3403 is located on the lane associated with the target trajectory L1. That is, using the positioning technique described above, the processor may determine the position of the host vehicle 3401. The image coordinates of the target vehicle 3403 may be mapped by line V1. As described above, image coordinates can be converted to map coordinates. Based on the map coordinates and the intersection of line V1 and lane L1, the processor may determine the position of the target vehicle 3403.

[0381] Alternatively, or further, the host vehicle may acquire two or more images of the target vehicle 3403 and observe how the representation of the target vehicle changes within the acquired images. For example, as the host vehicle 3401 and the target vehicle 3403 approach a merging point, the acquired images containing the target vehicle 3403 will show that the proportion of the acquired image occupied by the target vehicle 3403 increases as the distance between the host vehicle 3401 and the target vehicle 3403 decreases. Based on the change in the image representation size of the target vehicle 3403 (or based on any other observable image features related to the relative movement between the host vehicle 3401 and the target vehicle 3403), the processor may estimate the lane the target vehicle 3403 is traveling in. For example, the processor may access a stored sparse map and determine the possible lanes that exist on the opposite side of the obstruction 3405, as observed in the acquired images shown in Figure 34B. In the illustrated example, the processor may determine that there is one lane on the opposite side of the obstruction 3405 and that this lane is associated with the target trajectory L1. The processor may examine the observed motion characteristics associated with the target vehicle 3403 and determine whether these motion characteristics match the movement of the target vehicle along the target trajectory L1. If they do, the navigation processor of the host vehicle 3401 may estimate that the target vehicle 3403 is traveling along the path L1 in the lane on the opposite side of the obstruction 3405. With this information, the host vehicle processor may estimate the future path of the target vehicle. This estimation includes, for example, estimating how the target vehicle may interact with the host vehicle or how they may merge at an upcoming lane merge point. In some cases, even if the target vehicle is partially obscured, the host vehicle may slow down at an upcoming lane merge point to yield to the target vehicle. In some cases, the host vehicle may accelerate at a lane merge point to remain ahead of the target vehicle. The host vehicle determines the distance to the target vehicle based on sparse map information, and the image coordinates of the target vehicle 3403 may be mapped by line V1.

[0382] In the illustrated example, only one lane is shown, but the lane estimation and difference determination techniques can also be used when multiple lanes are not visible from the view of the host vehicle's image acquisition device. For example, if there are two, three, or more lanes that are not visible due to an obstruction 3405, the host vehicle processor can estimate which of the obscured lanes the target vehicle is traveling in by comparing the observed movement features of the target vehicle in two or more acquired images with the lane / trajectory information stored in the sparse map. From this information, the target vehicle's future path can be determined.

[0383] Figure 35A shows an example with multiple lanes that are not visible. Figure 35B shows the image coordinates of the target vehicle 3503 at t1 and t2 mapped by lines V1 and V2, respectively. The host vehicle 3501 is approaching an intersection with a two-lane road 3507. The processor can determine whether the identified target vehicle 3503 is in lane L1 or lane L2. This determination can be important for the appropriate navigation operation. For example, if the target vehicle is determined to be in the lane associated with trajectory L1, there may not be enough space for the host vehicle to turn left and enter the lane associated with trajectory L1. On the other hand, if the host vehicle processor determines, based on the observed motion characteristics of vehicle 3503 and estimation techniques based on stored map data, that the target vehicle 3503 is traveling in the lane associated with trajectory L2, the processor can conclude that there is enough space for the host vehicle 3501 to turn left and enter the lane associated with trajectory L1 without obstructing the detected target vehicle. In one embodiment, the processor may determine or estimate the speed of the target vehicle based on the distance from the host vehicle to the target vehicle and the position of the target vehicle in the image. For example, based on lines V1 and V2, the processor may determine the distance traveled by the target vehicle 3503 over a known time interval t2-t1. Thus, the speed of the target vehicle can be determined.

[0384] As shown in Figure 35B, if the host vehicle 3501 approaches an intersection between time t1 and t2, the distance from the host vehicle 3501 to the target vehicle 3503 decreases, and the size of the target vehicle 3503 in image 3510 also increases. In one embodiment, the processor may use the following equation to determine the distance Z1 from the host vehicle 3501 to the target vehicle 3503 at t1 and the distance Z2 from the host vehicle 3501 to the target vehicle 3503 at t2.

number

number

[0385] In the formula, w1 may represent the size of the target vehicle 3503 in image 3510 at t1, w2 may represent the size of the target vehicle 3503 in image 3510 at t2, and dZ may represent the distance traveled by the host vehicle 3501 between t1 and t2. This can be detected by the processor using the image analysis techniques described above. The dZ value can be determined from information from a speed sensor (e.g., a speedometer). Based on the calculated Z1 and Z2, the processor can determine the lane position of the target vehicle 3503.

[0386] Figure 36A shows another example of a multi-lane road where both the host vehicle 3601 and the target vehicle 3603 are approaching a lane merge point. Figure 36B shows the image coordinates of the target vehicle 3603 at times t1 and t2, mapped by V1 and V2, respectively. As shown in Figure 36A, the curved lane 3609 merges with road 3607. For the sake of simplicity, the processor may approximate the curved lane 3609 as a straight line 3613. The processor may use the following equations to determine the distance Z1 from the host vehicle 3601 to the target vehicle 3603 at t1 and the distance Z2 from the host vehicle 3601 to the target vehicle 3603 at t2.

number

number

[0387] In the formula, dZ may represent the distance traveled by the host vehicle 3601 between t1 and t2, a may represent the angle between line 3613 and road 3607, w1 may represent the size of the target vehicle 3603 in image 3610 at t1, and w2 may represent the size of the target vehicle 3603 in image 3610 at t2. In one embodiment, for simplicity, the processor may virtually rotate the image acquisition device (e.g., camera) so that the optical axis is perpendicular to lanes L1 and L2. This allows the movement of the target vehicle 3603 to be ignored. Furthermore, based on the calculated Z1 and Z2, the processor may determine the lane position of the target vehicle 3603 by estimation based on comparison with stored map data (e.g., determining which lane trajectory best matches the observed motion characteristics of the target vehicle).

[0388] In one embodiment, the processor may determine scale change information for an identified object. This scale change information may include size change information for the identified object within the image. For example, in image 3510, the processor may determine whether the size of the target vehicle 3503 (or an identifiable part or portion of the target vehicle) in the image is larger at time t2 than at time t1. The scale change information may further include position change information for the identified object within the image. For example, in image 3510, the processor may detect the movement of the target vehicle 2503 from right to left within the image between time t1 and t2. The size of an object within an image may be represented by pixels. In one embodiment, the time interval may be determined based on the image's timestamp. In one embodiment, the scale change information may include configuration information relating to the road's lane configuration. Based on image analysis, the processor may determine the road's lane configuration. The configuration may include lane features (e.g., straight lane, detour, 2-meter wide lane, forward right turn, etc.).

[0389] In one embodiment, as described above, the processor may determine a desired navigation action based on the determined lane position of the identified object. For example, if it is determined that the target vehicle 3503 is located in lane L2 in Figure 35A, the processor may cause the host vehicle 3501 to turn left and safely enter L1. If it is determined that the target vehicle 3503 is located in lane L1, the processor may cause the host vehicle 3501 to slow down and wait until it can turn left. In another example, if the target vehicle 3603 is located in lane L1 in Figure 36A, the processor may cause the host vehicle 3601 to slow down to avoid colliding with the target vehicle 3603. Furthermore, as described above, the processor may determine a navigation action based on the determined lane position of the target vehicle and the RSS safe distance.

[0390] Figure 37 is a flowchart of host vehicle navigation, consistent with the disclosed embodiments. In step 3701, the processor may receive at least one image representing the host vehicle environment from an image acquisition device. This image may be raw or processed data sent from image acquisition devices 122, 124, and / or 126 via a network or communication path. The data may include data described in some image format.

[0391] In step 3703, the processor may analyze at least one of several images to identify a first object in the host vehicle's environment. A second object in the vehicle's environment obscures at least part of the first object and the road on which the first object is located. As described above, images 3410, 3510, and 3610 may represent the host vehicle's environment. As described above, the processor may analyze the images and identify objects in the images using image analysis techniques, including, for example, object recognition, image segmentation, feature extraction, optical character recognition (OCR), object-based image analysis, shape domain techniques, edge detection methods, and pixel-based detection. For example, based on the analysis, the processor may identify that the first target vehicle 3503 is obscured by an obstruction 3505 in image 3510.

[0392] In step 3705, the processor may determine scale change information for the first object based on at least two of the multiple images. For example, the processor may compare image 3510, which contains images of the target vehicle 3503, in both t1 and t2, and determine the size change of the target vehicle 3503 in the images.

[0393] In step 3707, based on the scale change information determined for the first object, the processor may determine the lane position of the first object relative to the road lane on which the first object is located. As described above, based on Z1 and Z2, the processor may determine the lane position of the target vehicle.

[0394] In step 3709, the processor may determine navigation actions for the host vehicle based on the determined lane position of the first object. As described above, for example, if it is determined that the target vehicle 3503 is located in lane L2 in Figure 35A, the processor may cause the host vehicle 3501 to turn left and safely enter L1. If it is determined that the target vehicle 3503 is located in lane L1, the processor may cause the host vehicle 3501 to slow down and wait until it can turn left.

[0395] Determining the road location of the target vehicle based on its tracking trajectory.

[0396] As mentioned above, the target vehicle may be partially invisible in the acquired image. In such cases, the distance to the detected target vehicle can be determined by comparing it with the stored sparse map information. Alternatively, or further, as described above, the movement characteristics of the detected target vehicle (observed by image analysis) can be compared with the sparse map information to estimate the lane the target vehicle is traveling in, the expected direction of travel, etc.

[0397] As described above, in one embodiment, a processor for image processing may receive one or more acquired images representing the surrounding environment of a host vehicle. In one embodiment, as described above, the processor may perform image analysis on the set of images to identify objects within the set of images. The processor may identify objects in multiple images, multiple landmarks associated with road segments (e.g., road signs). Objects may be identified by any of the techniques disclosed above, such as object recognition, image segmentation, feature extraction, optical character recognition (OCR), object-based image analysis, shape region techniques, edge detection methods, and pixel-based detection.

[0398] In one embodiment, based on analysis, the processor may identify a first object in the environment (e.g., a target vehicle). The first object and the road on which the first object is located are partially obscured by a second object (e.g., an obstruction, a road sign, a roundabout, etc.). Figure 38A shows the position of the host vehicle 3801 relative to an object on a road segment. The processor of the host vehicle 3801 may detect the target vehicle 3803 and the roundabout 3805 by analyzing one or more acquired images. Based on the determined position relative to the host vehicle, the processor associated with the host vehicle may receive map information associated with the host vehicle's environment, as described above. For example, a sparse map associated with the road segment on which the host vehicle 3801 is traveling may be provided to the host vehicle 3801.

[0399] Figure 38A shows the positions of the host vehicle 3801 relative to the target vehicle 3803 and the roundabout 3805 at times t1, t2, and t3. Based on the position of the acquired images of the target vehicle 3803 and / or other movement features associated with the target vehicle (e.g., object scaling between acquired images), the processor can determine the trajectory of the target vehicle 3803 (e.g., the tracked trajectory traveled by the target vehicle between times t1 and t3). The target vehicle 3803, having entered the roundabout, may turn right. The initial turning angle may be approximately 90 degrees to the right (relative to the original direction of travel at times t1 and t2). Furthermore, as the target vehicle travels around the roundabout, it continues to turn left until it exits the roundabout. Figure 38B shows an exemplary composite image representing two different images received by the processor of the host vehicle 3801 from the image acquisition device at time t2 and the subsequent time t3. The combined images show the movement of the target vehicle 3803 from time t2 to time t3. The target vehicle 3803 is partially obscured by an obstruction 3805.

[0400] As described above, based on the images captured between t1 and r3, the processor can track changes in the size (or width (w)) of the target vehicle 3803 in the images. The processor can also track the direction of travel of the target vehicle. From this information, the travel path of the target vehicle can be determined, even if it is partially invisible. By comparing the determined travel path with the lane trajectories stored in the sparse map, it is possible to identify which of the stored lane trajectories matches the travel path of the target vehicle determined from the image analysis. By confirming the travel lane in this way, the host vehicle navigation processor can determine the expected travel path of the target vehicle in the preceding time and determine the navigation actions for the host vehicle regarding the travel lane determined for the detected target vehicle.

[0401] Figure 38C shows the relationship between the position (x) and width (w) of the target vehicle 3803 in the image. As shown in the figure, an inflection point may occur at the position where the target vehicle 3803 enters a roundabout. Based on the inflection point, the processor may select an image captured at the time the inflection point occurred. Based on the image analysis described above, the processor may determine the distance Z from the host vehicle 3801 to the target vehicle 3808. Based on the determined distance Z and the image width (w), the processor may calculate the actual vehicle width W using the formula described above. Once both the width (w) in the image and the actual width W have been determined, the processor may calculate the distance Z from the host vehicle 3801 to the target vehicle 3803 at any point where the image width (w) has been determined by the image analysis. The above method for determining the distance from the host vehicle to the target vehicle can be applied to any target vehicle trajectory where an inflection point occurs in the figure showing the relationship between the image width (w) and the image position (x).

[0402] Figure 39A shows the position of the host vehicle 3901 relative to the target vehicle 3903, the roundabout 3905, and the obstruction 3907 in an example. Figure 39B shows one image received by the processor of the host vehicle 3901 from the image acquisition device. In one embodiment, the processor may determine the lane position of the target vehicle 3903, i.e., whether the target vehicle 3903 is in the inner lane (T1) or the outer lane (T2). Based on the stored sparse map information, the processor may generate a tangent from the current position of the host vehicle to a known center position of the lane divider between lanes T1 and T2, as shown in Figure 39A. By analyzing the acquired image shown in Figure 39B using this tangent, the processor may determine the lane position of the target vehicle 3903. For example, if the target vehicle 3903 is to the left of the tangent in the image, the processor may determine that the target vehicle 3903 is in the outer lane (T2). If the target vehicle 3903 is to the right of the tangent in the image, the processor may determine that the target vehicle 3903 is in the inner lane (T1). The above method for determining the lane position of the target vehicle can be applied to any multi-lane curve or roundabout.

[0403] Figure 40A shows another example in which a host vehicle 4001 approaching a target vehicle 4003 is partially obscured by an obstruction 4007. Figure 40B shows an exemplary image received by the processor from an image acquisition device associated with the host vehicle. Without being inconsistent with the disclosed embodiments, the processor may receive an image 4010 captured by the image acquisition device of the host vehicle 4001. As shown in image 4010, the bottom of the target vehicle 4003 is obscured by the obstruction 4007. Based on analysis of the acquired image, the processor may determine that the target vehicle 4003 is located along line V1 relative to the host vehicle, even though its exact position along line V1 is unknown. To determine the actual position of the target vehicle 4003, the processor may assume a distance (e.g., Z1 or Z2) from the host vehicle 4001 to the target vehicle 4003. As described above, the processor monitors the observed movement characteristics of the target vehicle 4003 between two or more captured images and can determine the observed trajectory of the target vehicle for each hypothetical distance (e.g., Z1 or Z2) that may be associated with different observed trajectories of the target vehicle. The processor can compare the observed trajectory for the target vehicle with stored map information showing drivable lanes and their associated trajectories on the opposite side of the obstruction 4007. This comparison allows the processor to determine which of the hypothetical trajectories is most likely to be the target vehicle (e.g., the one that best matches the actual lane and trajectory values ​​represented in the stored map). In some cases, the processor can rank each hypothetical trajectory according to its confidence (or other numerical value) in matching the actual trajectory of the target vehicle (e.g., which observed trajectory best matches an available or relevant mapping trajectory).

[0404] For example, as shown in Figure 40A, if the actual position of the target vehicle is T1, the...

Claims

1. A system for navigating a host vehicle, wherein the system A processing device comprising at least one having a circuit and memory, When executed by the circuit, the memory is sent to the at least one processing device. Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Having an order to carry out, Comparing the determined trajectory with the plurality of target trajectories includes comparing one or more inflection points associated with the determined trajectory with one or more inflection points associated with the plurality of trajectories. system.

2. A system for navigating a host vehicle, wherein the system is A processing device comprising at least one having a circuit and memory, When executed by the circuit, the memory is sent to the at least one processing device. Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Having an order to carry out, The first estimated position and the second estimated position are determined based on the size of the representation of the target vehicle in the plurality of images. system.

3. The system according to claim 1 or 2, wherein the first estimated position and the second estimated position are determined based on the estimated distance between the host vehicle and the target vehicle.

4. The system according to any one of claims 1 to 3, further comprising comparing the determined trajectory with the plurality of target trajectories, and ranking the first trajectory and the second trajectory based on their degree of agreement with the determined trajectory.

5. The system according to claim 4, wherein the target trajectory is selected from the first trajectory and the second trajectory based on the ranking.

6. The system according to any one of claims 1 to 5, wherein comparing the determined trajectory with the plurality of target trajectories includes applying a trained machine learning model.

7. The system according to any one of claims 1 to 6, wherein determining the trajectory of the target vehicle within the aforementioned period is further based on the output of at least one navigation sensor of the host vehicle.

8. The system according to claim 7, wherein the at least one navigation sensor includes a GPS device, a speed sensor, or an accelerometer.

9. The system according to any one of claims 1 to 8, wherein determining the position of the target vehicle relative to the road is further based on GPS information.

10. The system according to any one of claims 1 to 9, wherein the navigation operation includes causing the navigation actuator of the host vehicle to perform adjustment.

11. The system according to claim 10, wherein the navigation actuator includes at least one of the steering mechanism of the host vehicle, a brake, and an accelerator.

12. The system according to any one of claims 1 to 11, wherein the position of the target vehicle occurs on a curve in the road.

13. The system according to any one of claims 1 to 12, wherein the position of the target vehicle occurs at a roundabout of the road.

14. A method for navigating the host vehicle, Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Equipped with, Comparing the determined trajectory with the plurality of target trajectories includes comparing one or more inflection points associated with the determined trajectory with one or more inflection points associated with the plurality of trajectories. method.

15. A method for navigating a host vehicle, Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Equipped with, The first estimated position and the second estimated position are determined based on the size of the representation of the target vehicle in the plurality of images. method.

16. The method according to claim 14 or 15, wherein the first estimated position and the second estimated position are determined based on the estimated distance between the host vehicle and the target vehicle.

17. The method according to any one of claims 14 to 16, further comprising comparing the determined trajectory with the plurality of target trajectories, and ranking the first trajectory and the second trajectory based on their degree of agreement with the determined trajectory.

18. A program that causes at least one processor to execute a method for navigating a host vehicle, wherein the method is Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Equipped with, Comparing the determined trajectory with the plurality of target trajectories includes comparing one or more inflection points associated with the determined trajectory with one or more inflection points associated with the plurality of trajectories. program.

19. A program that causes at least one processor to perform a method for navigating a host vehicle, wherein the method is: Receiving multiple images showing the environment of the host vehicle, which were captured by the image acquisition device within the specified period, To identify the target vehicle in the environment of the host vehicle, at least one of the plurality of images is analyzed. Receiving map information including a plurality of target trajectories associated with the environment of the host vehicle, wherein the plurality of target trajectories include at least a first trajectory associated with a first driving lane along a road segment and a second trajectory associated with a second driving lane along the road segment. Based on the analysis of the plurality of images, the first estimated position of the target vehicle at the first time and the second estimated position of the target vehicle at the second time are determined, and the determination is made that the first time and the second time fall within the specified period. Based on the first estimated position and the second estimated position, the trajectory of the target vehicle within the period is determined, The determined trajectory is compared with the plurality of target trajectories to identify one target trajectory from the plurality of target trajectories through which the target vehicle passes, wherein the target trajectory includes the first trajectory or the second trajectory. Based on the identified target trajectory, the position of the host vehicle relative to the road in the environment is determined. Based on the determined position of the target vehicle, a navigation operation for the host vehicle is determined. Equipped with, The first estimated position and the second estimated position are determined based on at least one of the size of the representation of the target vehicle in the plurality of images, or the estimated distance between the host vehicle and the target vehicle. program.

Images