System for vehicle navigation based on image analysis

Abstract

A system and method for vehicle navigation are provided. In one implementation, at least one processing device may receive at least one captured image representing an environment of the host vehicle from a camera of the host vehicle. The processing device may analyze one or more pixels of the at least one captured image to determine whether the one or more pixels represent at least a portion of a target vehicle. For pixels determined to represent at least a portion of the target vehicle, the processing device may determine one or more estimated distance values ​​from the one or more pixels to at least one edge of a surface of the target vehicle, and generate at least a portion of a boundary for the target vehicle based on the analysis of the one or more pixels, including the determined one or more distance values ​​associated with the one or more pixels.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims the benefit of priority of U.S. Provisional Patent Application No. 62,852,761, filed on May 24, 2019; U.S. Provisional Patent Application No. 62 / 957,009, filed on January 3, 2020; and U.S. Provisional Patent Application No. 62 / 976,059, filed on February 13, 2020. The above applications are hereby incorporated by reference in their entireties.

[0002] This disclosure generally relates to autonomous vehicle navigation.

Background Art

[0003] As technology continues to evolve, the goal of fully autonomous vehicles that can navigate on roads is becoming more realistic. Autonomous vehicles need to consider various factors and make appropriate judgments to safely and accurately reach the intended destination based on those factors. 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., GPS devices, speed sensors, accelerometers, suspension sensors, etc.). At the same time, to navigate to the destination, autonomous vehicles need to identify their position within a specific road (e.g., a specific lane within a multi-lane road), navigate alongside other vehicles, avoid obstacles and pedestrians, observe traffic signals and signs, and move from one road to another at appropriate intersections or interchanges. Interpreting the vast amount of information collected by autonomous vehicles as they travel to the destination poses many design challenges. The vast amount of data (e.g., captured image data, map data, GPS data, sensor data, etc.) that autonomous vehicles may need to analyze, access, and / or store presents issues that can actually limit or negatively impact autonomous navigation. Additionally, if autonomous vehicles rely on conventional mapping technologies to navigate, the vast amount of data required to store and update the maps presents difficult challenges. SUMMARY OF THE INVENTION

[0004] Embodiments according to the present disclosure provide systems and methods for autonomous vehicle navigation. The disclosed embodiments may use cameras to provide autonomous vehicle navigation features. For example, according to embodiments of the present disclosure, the disclosed system may include one, two, or more than two cameras for monitoring the environment of the vehicle. The disclosed system may provide a navigation response, for example, based on the analysis of images captured by one or more of the cameras.

[0005] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may receive at least one captured image representing the environment of the host vehicle from a camera of the host vehicle, and analyze one or more pixels of the at least one captured image to determine whether one or more pixels represent at least a portion of a target vehicle. In the case of pixels determined to represent at least a portion of the target vehicle, the processor may determine one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle. Further, the processor may generate at least a portion of a boundary for the target vehicle based on the analysis of the one or more pixels, including the determined one or more distance values associated with the one or more pixels.

[0006] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may receive at least one captured image representing the environment of the host vehicle from a camera of the host vehicle, and analyze one or more pixels of the at least one captured image to determine whether one or more pixels represent a target vehicle, where at least a portion of the target vehicle is not represented within the at least one captured image. The processor may be further configured to determine an estimated distance from the host vehicle to the target vehicle, the estimated distance being at least partially based on the portion of the target vehicle that is not represented within the at least one captured image.

[0007] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may receive at least one captured image representing the environment of the host vehicle from a camera of the host vehicle, and analyze two or more pixels of the at least one captured image to determine whether the two or more pixels represent at least a portion of a first target vehicle and at least a portion of a second target vehicle. The processor may be programmed to determine that a portion of the second target vehicle is included in the representation of a reflection on the surface of the first target vehicle. Based on the analysis of the two or more pixels and the determination that a portion of the second target vehicle is included in the representation of a reflection on the surface of the first target vehicle, the processor may be further programmed to generate at least a portion of a boundary for the first target vehicle and not generate a boundary for the second target vehicle.

[0008] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may receive at least one captured image representing the environment of the host vehicle from a camera of the host vehicle, and analyze two or more pixels of the at least one captured image to determine whether the two or more pixels represent at least a portion of a first target vehicle and at least a portion of a second target vehicle. The processor may be programmed to determine whether the second target vehicle is being carried or towed by the first target vehicle. Based on the analysis of the two or more pixels and the determination of whether the second target vehicle is being carried or towed by the first target vehicle, the processor may be further programmed to generate at least a portion of a boundary for the first target vehicle and not generate a boundary for the second target vehicle.

[0009] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may be programmed to receive, from a camera of the host vehicle, a first captured image representative of the environment of the host vehicle, and to analyze one or more pixels of the first captured image to determine whether the one or more pixels represent at least a portion of a target vehicle. For pixels determined to represent at least a portion of the target vehicle, the processor may determine one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle. The processor may be programmed to generate at least a portion of a first boundary for the target vehicle based on an analysis of one or more pixels of the first captured image, including the determined one or more distance values associated with the one or more pixels of the first captured image. The processor may be further programmed to receive, from a camera of the host vehicle, a second captured image representative of the environment of the host vehicle, and to analyze one or more pixels of the second captured image to determine whether the one or more pixels represent at least a portion of a target vehicle. For pixels determined to represent at least a portion of the target vehicle, the processor may determine one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle. The processor may be programmed to generate at least a portion of a second boundary for the target vehicle based on an analysis of one or more pixels of the second captured image, including the determined one or more distance values associated with the one or more pixels of the second captured image, and based on the first boundary.

[0010] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may be programmed to receive from a camera of the host vehicle two or more images captured from the environment of the host vehicle, and analyze the two or more images to identify at least a partial representation of a first object and at least a partial representation of a second object. The processor may determine a first region of at least one image associated with the first object and the type of the first object, and may determine a second region of at least one image associated with the second object and the type of the second object, wherein the type of the first object is different from the type of the second object.

[0011] In one embodiment, a navigation system for a host vehicle may comprise at least one processor. The processor may be programmed to receive from a camera of the host vehicle at least one image captured from the environment of the host vehicle, and analyze the at least one image to identify at least a partial representation of a first object and at least a partial representation of a second object. The processor may be programmed to determine, based on the analysis, at least one aspect of the geometry of the first object and at least one aspect of the geometry of the second object. Further, the processor may be programmed to generate a first label associated with a region of at least one image that includes a representation of the first object, based on at least one aspect of the geometry of the first object, and generate a second label associated with a region of at least one image that includes a representation of the second object, based on at least one aspect of the geometry of the second object.

[0012] According to other disclosed embodiments, a non-transitory computer-readable storage medium may store program instructions that are executable by at least one processing device and that perform any of the methods described herein.

[0013] The foregoing summary and the following detailed description are merely illustrative and explanatory and are not restrictive of the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate various embodiments disclosed herein.

[0015]

Figure 1

[0016]

Figure 2A

[0017]

Figure 2B

[0018]

Figure 2C

[0019]

Figure 2D

[0020]

Figure 2E

[0021]

Figure 2F

[0022]

Figure 3A

[0023]

Figure 3B

[0024]

Figure 3C

[0025]

Figure 3D

[0026]

Figure 4

[0027]

Figure 5A

[0028]

Figure 5B

[0029]

Figure 5C

[0030]

Figure 5D

[0031]

Figure 5E

[0032]

Figure 5F

[0033]

Figure 6

[0034]

Figure 7

[0035]

Figure 8

[0036]

Figure 9A

[0037]

Figure 9B

[0038]

Figure 10

[0039]

Figure 11A

[0040]

Figure 11B

Figure 11C

[0041]

Figure 11D

[0042]

Figure 12

[0043]

Figure 13

[0044]

Figure 14

[0045]

Figure 15

[0046]

Figure 16

[0047]

Figure 17

[0048]

Figure 18

[0049]

Figure 19

[0050]

Figure 20

[0051]

Figure 21

[0052]

Figure 22

[0053]

Figure 23

[0054]

Figure 24A

Figure 24B

Figure 24C

Figure 24D

[0055]

Figure 24E

[0056]

Figure 24F

[0057]

Figure 25A

[0058]

Figure 25B

[0059]

Figure 26A

[0060]

Figure 26B

[0061]

Figure 27

[0062]

Figure 28

[0063]

Figure 29

[0064]

Figure 30A

Figure 30B

[0065]

Figure 31A

[0066]

Figure 31B

[0067]

Figure 32A

[0068]

Figure 32B

[0069]

Figure 33

[0070]

Figure 34A

Figure 34B

[0071]

Figure 35

[0072]

Figure 36

[0073]

Figure 37

Mode for Carrying Out the Invention

[0074] The following detailed description refers to the accompanying drawings. Whenever possible, the same reference numbers are used in the drawings and the following description to refer to the same or similar parts. Some exemplary embodiments are described herein, but modifications, adaptations, and other implementations are possible. For example, substitutions, additions, or changes can be made to the components shown in the drawings, and the exemplary methods described herein can be changed by substituting, reordering, deleting, or adding steps of the disclosed methods. Therefore, the following detailed description is not limited to the disclosed embodiments and examples. Instead, the appropriate scope is defined by the appended claims.

[0075] Overview of Autonomous Vehicles

[0076] When used throughout this disclosure, the term "autonomous vehicle" refers to a vehicle that can perform at least one navigation change without driver input. A "navigation change" refers to one or more changes to the steering, braking, or acceleration of the vehicle. To be autonomous, the vehicle does not need to be fully automatic (e.g., operate completely without a driver or driver input). Rather, an autonomous vehicle includes a vehicle that can operate under driver control during certain time periods and without driver control during other time periods. An autonomous vehicle can include a vehicle that controls only some aspects of vehicle navigation, such as steering (e.g., to maintain the vehicle course between vehicle lane constraints), but can leave other aspects (e.g., braking) to the driver. In some cases, an autonomous vehicle can handle some or all aspects of vehicle braking, speed control, and / or steering.

[0077] Since human drivers typically rely on visual cues and observations to control a vehicle, the transportation infrastructure is built accordingly, and lane markings, traffic signs, and traffic lights are designed to provide all visual information to the driver. In view of these design features of the transportation infrastructure, an autonomous vehicle can include a camera and a processing unit that analyzes visual information captured from the vehicle's environment. The visual information can include, for example, components of the transportation infrastructure observable by a driver (e.g., lane markings, traffic signs, traffic lights, etc.) and other obstacles (e.g., other vehicles, pedestrians, debris, etc.). Additionally, an autonomous vehicle can also use stored information, such as information that provides a model of the vehicle's environment when navigating. For example, the vehicle can 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 vehicle's environment while the vehicle is in motion, and the vehicle (and other vehicles) can use the information to identify its position in the model.

[0078] In some embodiments of the present disclosure, an autonomous vehicle may use information obtained during navigation (e.g., from a camera, GPS device, accelerometer, speed sensor, suspension sensor, etc.). In other embodiments, an autonomous vehicle may use information obtained from past navigation by the vehicle (or other vehicles) during navigation. In yet other embodiments, an autonomous vehicle may use a combination of information obtained during navigation and information obtained from past navigation. The following sections provide an overview of the system according to the disclosed embodiments, followed by an overview of the forward imaging system and method according to that system. The following sections disclose systems and methods for constructing, using, and updating a sparse map for autonomous vehicle navigation.

[0079] System Overview

[0080] FIG. 1 is a block diagram representation of a system 100 according to an exemplary disclosed embodiment. The system 100 may include various components depending on specific implementation requirements. In some embodiments, the 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 capture devices (e.g., cameras) such as image capture device 122, image capture device 124, image capture device 126, etc. The system 100 may also include a data interface 128 that communicatively connects the processing device 110 to the image acquisition device 120. For example, the data interface 128 may include one or more arbitrary wired links and / or wireless links for transmitting image data acquired by the image acquisition device 120 to the processing unit 110.

[0081] The wireless transceiver 172 may include one or more devices configured to exchange transmissions via a wireless interface using radio frequency, infrared frequency, magnetic field, or electric field with one or more networks (e.g., cellular, Internet, etc.). 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 communications from the host vehicle to one or more remotely located servers. Such transmissions may also include (unidirectional or bidirectional) communications between the host vehicle and one or more target vehicles within the host vehicle's environment (e.g., taking into account or along with a target vehicle within the host vehicle's environment to facilitate adjustment of the host vehicle's navigation), and further broadcast transmissions to unspecified recipients in the vicinity of the transmitting vehicle.

[0082] 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 pre-processor (such as an image pre-processor), a graphics processing unit (GPU), a central processing unit (CPU), support circuitry, a digital signal processor, an integrated circuit, a memory, or any other type of device suitable for executing an application and processing and analyzing images. In some embodiments, the application processor 180 and / or the image processor 190 may include any type of single-core or multi-core processor, a mobile device microcontroller, a central processing unit, etc. For example, various processing devices are available, including processors available from manufacturers such as Intel®, AMD®, etc., or GPUs available from manufacturers such as NVIDIA®, ATI®, etc., and may include various architectures (e.g., x86 processors, ARM®, etc.).

[0083] In some embodiments, application processor 180 and / or image processor 190 may include any EyeQ series processor available from Mobileye®. These processor designs each include multiple processing units with local memory and instruction sets. Such a processor may include video input that receives image data from multiple image sensors and may also include a video output function. In one example, EyeQ2® operates at 332 MHz and uses 90nm-micron technology. The EyeQ2® architecture consists of two floating point hyperthreaded 32-bit RISC CPUs (MIPS32® 34K® cores), five vision calculation engines (VCEs), three vector microcode processors (VMP®), a Denali 64-bit mobile DDR controller, a 128-bit internal acoustic interconnect, a dual 16-bit video input and an 18-bit video output controller, 16-channel DMA, and several peripherals. The MIPS34K CPU manages the five VCEs, three VMPs (trademarks), and DMA, a second MIPS34K CPU, and multi-channel DMA, and other peripherals. The five VCEs, three VMPs®, and MIPS34K CPU can perform intensive vision calculations required by multifunctional bundle applications. In another example, in the disclosed embodiments, EyeQ3®, a third-generation processor that is more than six times more powerful than EyeQ2®, may be used. In other examples, EyeQ4® and / or EyeQ5® may be used in the disclosed embodiments. Of course, newer or future EyeQ processing devices may be used with the disclosed embodiments.

[0084] Any of the processing devices disclosed in this specification can be configured to perform a specific function. Configuring a processing device such as the described EyeQ processor or any other controller or microprocessor to perform a specific function can include programming computer-executable instructions and providing those instructions to the processing device for execution during the operation of the processing device. In some embodiments, configuring the processing device can include directly programming architectural instructions into the processing device. For example, processing devices such as field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), etc. can be configured using, for example, one or more hardware description languages (HDLs).

[0085] In other embodiments, configuring the processing device can include storing executable instructions on a memory accessible by the processing device during operation. For example, the processing device can access the memory during operation to obtain and execute the stored instructions. In any case, the processing device configured to perform the detection, image analysis, and / or navigation functions disclosed herein represents a dedicated hardware-based system that controls multiple hardware-based components of the host vehicle.

[0086] FIG. 1 shows two separate processing devices included in processing unit 110, although more or fewer processing devices can also be used. For example, in some embodiments, a single processing device can be used to achieve the tasks of application processor 180 and image processor 190. In other embodiments, these tasks can be performed by three or more processing devices. Further, in some embodiments, system 100 can include one or more of processing unit 110 without including other components such as image acquisition unit 120.

[0087] The processing unit 110 may include various types of devices. For example, the processing unit 110 may include various devices such as a controller, an image pre-processor, a central processing unit (CPU), a graphics processing unit (GPU), support circuits, a digital signal processor, an integrated circuit, a memory, or any other type of device that processes and analyzes images. The image pre-processor may include a video processor that captures, digitizes, and processes images from an image sensor. The CPU may include any number of microcontrollers or microprocessors. The GPU may also include any number of microcontrollers or microprocessors. The support circuits may be any number of circuits generally 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 a database and image processing software. The memory may include any number of random access memories, read-only memories, flash memories, disk drives, optical storage devices, tape storage devices, removable storage devices, 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 with the processing unit 110.

[0088] Each of the memories 140, 150 may contain software instructions that, when executed by a processor (e.g., application processor 180 and / or image processor 190), can control the operation of various aspects of the system 100. These memory units may include various databases and image processing software as well as trained systems such as, for example, neural networks or deep neural networks. The memory units may include random access memory (RAM), read only memory (ROM), flash memory, disk drives, optical storage devices, tape storage devices, removable storage devices, 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 with the application processor 180 and / or the image processor 190.

[0089] The position sensor 130 may include any type of device suitable for identifying 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 identify the user's position and speed by processing signals broadcast by Global Positioning System satellites. The position information from the position sensor 130 may be provided to the application processor 180 and / or the image processor 190.

[0090] In some embodiments, the system 100 may include components such as a speed sensor (e.g., a tachometer, a speedometer) for measuring the speed of the vehicle 200 and / or an accelerometer (either single-axis or multi-axis) for measuring the acceleration of the vehicle 200.

[0091] The user interface 170 may include any device suitable for providing information or receiving input from one or more users of the system 100. In some embodiments, the user interface 170 may include, for example, a touch screen, a microphone, a keyboard, a pointer device, a track wheel, a camera, a knob, buttons, etc., and may include a user input device. Using such input devices, the user can type commands or information, provide voice commands, use buttons, pointers or eye-tracking functions, or select menu options on the screen through any other suitable technique for communicating information to the system 100, thereby providing information input or commands to the system 100.

[0092] The user interface 170 may include one or more processing devices configured to provide information to the user or receive information from the user and process that information for use, for example, by the application processor 180. In some embodiments, such processing devices may execute instructions to recognize and track eye movements, receive and interpret voice commands, recognize and interpret touches and / or gestures made on a touch screen, respond to keyboard inputs or menu selections, and so on. In some embodiments, the user interface 170 may include a display, a speaker, a tactile device, and / or any other device for providing output information to the user.

[0093] The map database 160 can include any type of database that stores map data useful to the system 100. In some embodiments, the map database 160 can include data related to the positions of various items, such as roads, water features, geographical features, businesses, points of interest, restaurants, gas stations, etc., in a reference coordinate system. The map database 160 can store not only the positions of such items but also descriptors related to those items, including, for example, names associated with any of the stored features. In some embodiments, the map database 160 can be physically located with other components of the system 100. Alternatively or additionally, the map database 160 or a portion thereof can be located remotely with respect to other components of the system 100 (e.g., the processing unit 110). In such embodiments, information from the map database 160 can be downloaded to the network via a wired or wireless data connection (e.g., via a cellular network and / or the Internet, etc.). In some cases, the map database 160 can store a sparse data model that includes specific road features (e.g., lane markings) or a polynomial representation of the target trajectory of the host vehicle. Systems and methods for generating such maps will be discussed below with reference to FIGS. 8-19.

[0094] The image capture devices 122, 124, and 126 can each include any type of device suitable for capturing at least one image from the environment. Further, any number of image capture devices can be used to obtain images input to the image processor. Some embodiments can include only a single image capture device, while other embodiments can include two, three, or even four or more image capture devices. The image capture devices 122, 124, and 126 will be further described below with reference to FIGS. 2B-2E.

[0095] System 100 or various components of system 100 can be incorporated into various different platforms. In some embodiments, system 100 can be included in vehicle 200 as shown in FIG. 2A. For example, vehicle 200 can include processing unit 110 and any other components of system 100 as described above with respect to FIG. 1. In some embodiments, vehicle 200 can include only a single image capture device (e.g., a camera), and in other embodiments such as the embodiments discussed in connection with FIGS. 2B-2E, multiple image capture devices can be used. For example, as shown in FIG. 2A, either of image capture devices 122 and 124 of vehicle 200 can be part of an ADAS (Advanced Driver Assistance System) imaging set.

[0096] The image capture device included in vehicle 200 as part of image acquisition unit 120 can be positioned at any suitable location. In some embodiments, as shown in FIGS. 2A-2E and FIGS. 3A-3C, image capture device 122 can be positioned near the rearview mirror. This location can provide a line of sight similar to that of the driver of vehicle 200 and can assist the driver in determining what can be seen and what cannot be seen. Image capture device 122 can be located at any position near the rearview mirror, but placing image capture device 122 on the driver's side of the mirror can further assist in obtaining an image representing the driver's field of view and / or line of sight.

[0097] Other positions can also be used for the image capture device of the image acquisition unit 120. For example, the image capture device 124 can be disposed on or within the bumper of the vehicle 200. Such a position may be particularly suitable for an image capture device having a wide field of view. The line of sight of the image capture device disposed on the bumper can be different from that of the driver, and thus, the bumper image capture device and the driver do not always see the same object. The image capture devices (e.g., the image capture devices 122, 124, and 126) can also be disposed at other positions. For example, the image capture device can be disposed on 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, in the trunk of the vehicle 200, on the side of the vehicle 200, can be mounted on, positioned behind, or positioned in front of any window of the vehicle 200, and can be mounted in or near the front and / or rear lights of the vehicle 200, etc.

[0098] In addition to the image capture device, the vehicle 200 can include various other components of the system 100. For example, the processing unit 110 can be integrated into the engine control unit (ECU) of the vehicle or can be included in the vehicle 200 separately from the ECU. The vehicle 200 can also be provided with a position sensor 130 such as a GPS receiver, and the vehicle 200 can include a map database 160 as well as memory units 140 and 150.

[0099] As described above, the wireless transceiver 172 can transmit and / or receive data via one or more networks (e.g., cellular network, Internet, etc.). For example, the wireless transceiver 172 can upload the data collected by the system 100 to one or more servers and download data from one or more servers. Through the wireless transceiver 172, the system 100 can receive updates to the data stored in, for example, the map database 160, the memory 140, and / or the memory 150 periodically or when needed. Similarly, the wireless transceiver 172 can upload any data from the system 100 (e.g., images captured by the image acquisition unit 120, data received by the position sensor 130, other sensors, or the vehicle control system, etc.) and / or any data processed by the processing unit 110 to one or more servers.

[0100] The system 100 can upload data to a server (e.g., cloud) based on the privacy level setting. For example, the system 100 can implement a privacy level setting that regulates or restricts the type of data (including metadata) that can uniquely identify the vehicle and / or the driver / owner of the vehicle and is transmitted to the server. Such settings can be set by the user via the wireless transceiver 172, initialized by factory default settings, or set by the data received by the wireless transceiver 172.

[0101] In some embodiments, the system 100 can upload data according to a "high" privacy level. Under the setting of the setting, the system 100 can transmit data without any details about a specific vehicle and / or driver / owner (e.g., location information related to the route, captured images, etc.). For example, when uploading data according to the "high" privacy level, the system 100 can transmit data such as captured images and / or limited location information related to the route without including the vehicle identification number (VIN) or the name of the driver or owner of the vehicle.

[0102] Other privacy levels are intended. For example, the 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 manufacturer and / or model of the vehicle and / or vehicle type (e.g., passenger car, sports utility vehicle, truck, etc.). In some embodiments, the system 100 may upload data according to a "low" privacy level. Under the "low" privacy level setting, the system 100 may upload data sufficient to uniquely identify a particular vehicle, owner / driver, and / or part or all of the route traveled by the vehicle, and may include such information. Such "low" privacy level data may include, for example, one or more of a VIN, driver / owner name, starting point of the vehicle before departure, intended destination of the vehicle, manufacturer and / or model of the vehicle, type of the vehicle, etc.

[0103] FIG. 2A is a side view representation of an exemplary vehicle imaging system according to the disclosed embodiment. FIG. 2B is a top view representation of the embodiment shown in FIG. 2A. As shown in FIG. 2B, the disclosed embodiment may include a vehicle 200 having a system 100 incorporated therein that includes a first image capture device 122 positioned near the rearview mirror and / or near the driver of the vehicle 200, a second image capture device 124 positioned on or within a bumper region of the vehicle 200 (e.g., one of the bumper regions 210), and a processing unit 110.

[0104] As shown in FIG. 2C, both the image capture devices 122 and 124 may be positioned near the rearview mirror of the vehicle 200 and / or near the driver. Further, although two image capture devices 122 and 124 are shown in FIGS. 2B and 2C, it should be understood that other embodiments may include three or more image capture devices. For example, in the embodiments shown in FIGS. 2D and 2E, a first image capture device 122, a second image capture device 124, and a third image capture device 126 are included in the system 100 of the vehicle 200.

[0105] As shown in FIG. 2D, the image capture device 122 may be positioned near the rearview mirror of the vehicle 200 and / or near the driver, and the image capture devices 124 and 126 may be positioned on the bumper region of the vehicle 200 (e.g., one of the bumper regions 210). Also, as shown in FIG. 2E, the image capture devices 122, 124, and 126 may be positioned near the rearview mirror of the vehicle 200 and / or near the driver's seat. The disclosed embodiments are not limited to any particular number and configuration of image capture devices, and the image capture devices may be positioned at any suitable location within and / or on the vehicle 200.

[0106] It should be understood that the disclosed embodiments are not limited to vehicles and are applicable in other situations. It should also be understood that the disclosed embodiments are not limited to a particular type of vehicle 200 and may be applicable to all types of vehicles, including automobiles, trucks, trailers, and other types of vehicles.

[0107] The first image capture device 122 may include any suitable type of image capture device. The image capture device 122 may include an optical axis. In one example, the image capture device 122 may include an Aptina M9V024 WVGA sensor having a global shutter. In other embodiments, the image capture device 122 may provide a resolution of 1280×960 pixels and may include a rolling shutter. The image capture device 122 may include various optical elements. In some embodiments, one or more lenses may be included to provide, for example, a desired focal length and field of view of the image capture device. In some embodiments, a 6 mm lens or a 12 mm lens may be associated with the image capture device 122. In some embodiments, the image capture device 122 may be configured to capture an image having a desired field of view (FOV) 202 as shown in FIG. 2D. For example, the image capture device 122 may be configured to have a normal FOV within a range of 40 degrees to 56 degrees, including degrees greater than 46 degrees FOV, 50 degrees FOV, 52 degrees FOV, or 52 degrees FOV. Alternatively, the image capture device 122 may be configured to have a narrow FOV in the range of 23 to 40 degrees, such as 28 degrees FOV or 36 degrees FOV. Additionally, the image capture device 122 may be configured to have a wide FOV in the range of 100 to 180 degrees. In some embodiments, the image capture device 122 may include a wide-angle bumper camera or a bumper camera having a FOV of up to 180 degrees. In some embodiments, the image capture device 122 may be a 7.2M pixel image capture device having a horizontal FOV of about 100 degrees and an aspect ratio of about 2:1 (e.g., H×V = 3800×1900 pixels). Such an image capture device may be used instead of a three-dimensional image capture device configuration. Due to large lens distortion, the vertical FOV of such an image capture device may be much lower than 50 degrees in an implementation where the image capture device uses a radially symmetric lens. For example, such a lens is not radially symmetric, thereby allowing a vertical FOV greater than 50 degrees with a horizontal FOV of 100 degrees.

[0108] The first image capture device 122 may obtain a plurality of first images of a scene associated with the vehicle 200. The plurality of first images may each be obtained as a series of image scan lines, which may be captured using a rolling shutter. Each scan line may include a plurality of pixels.

[0109] The first image capture device 122 may have a scan rate associated with the acquisition of each of the first series of image scan lines. The scan rate may refer to the rate at which the image sensor can acquire the image data associated with each pixel included in a particular scan line.

[0110] The image capture devices 122, 124, and 126 may include any suitable type and number of image sensors, including, for example, a CCD sensor or a CMOS sensor. In one embodiment, the CMOS image sensor may be utilized with a rolling shutter, whereby each pixel within a row is read one at a time, and the scanning of the row proceeds row by row until the entire image frame is captured. In some embodiments, the rows may be sequentially captured from top to bottom with respect to the frame.

[0111] In some embodiments, one or more of the image capture devices disclosed herein (e.g., image capture devices 122, 124, and 126) may constitute a high-resolution imager and may have a resolution greater than 5M pixels, greater than 7M pixels, greater than 10M pixels, or more.

[0112] By using a rolling shutter, pixels in different rows can be exposed and captured at different times, whereby skew and other image artifacts can occur in the captured image frame. On the other hand, if the image capture device 122 is configured to operate using a global or synchronous shutter, all pixels can be exposed during a common exposure period for the same amount of time. As a result, the image data in the frame collected from a system that utilizes a global shutter represents a snapshot of the entire FOV (such as FOV202) at a particular time. In contrast, when applying a rolling shutter, each row in the frame is exposed and the data is captured at different times. Therefore, a moving object may appear distorted in an image capture device having a rolling shutter. This phenomenon will be described in more detail below.

[0113] The second image capture device 124 and the third image capture device 126 can be any type of image capture device. Similar to the first image capture device 122, each of the image capture devices 124 and 126 can include an optical axis. In one embodiment, each of the image capture devices 124 and 126 can include an Aptina M9V024 WVGA sensor having a global shutter. Alternatively, each of the image capture devices 124 and 126 can include a rolling shutter. Similar to the image capture device 122, the image capture devices 124 and 126 can be configured to include various lenses and optical elements. In some embodiments, the lenses associated with the image capture devices 124 and 126 can be the same as the FOV (such as FOV202) associated with the image capture device 122 or can provide a narrower FOV (such as FOV204 and 206). For example, the image capture devices 124 and 126 can have a FOV of 40 degrees, 30 degrees, 26 degrees, 23 degrees, 20 degrees, or less than 20 degrees.

[0114] Image capture devices 124 and 126 may acquire a plurality of second and third images of a scene associated with 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 may be captured using a rolling shutter. Each scan line or row may have a plurality of pixels. Image capture devices 124 and 126 may have second and third scan rates associated with the acquisition of each image scan line included within the second and third series.

[0115] Each of image capture devices 122, 124, and 126 may be positioned at any suitable location and in any suitable orientation with respect to vehicle 200. The relative positions of image capture devices 122, 124, and 126 may be selected to assist in fusing together information obtained from the image capture devices. For example, in some embodiments, the FOV (FOV204) associated with image capture device 124 may overlap partially or completely with the FOV (such as FOV202) associated with image capture device 122 and the FOV (such as FOV206) associated with image capture device 126.

[0116] Image capture devices 122, 124, and 126 may be disposed on vehicle 200 at any suitable relative height. In one example, there may be a height difference between image capture devices 122, 124, and 126, and the height difference may provide sufficient parallax information to enable stereoscopic analysis. For example, as shown in FIG. 2A, two image capture devices 122 and 124 are at different heights. There may also be a lateral displacement difference between image capture devices 122, 124, and 126, which may provide additional parallax information for stereoscopic analysis by, for example, processing unit 110. The lateral displacement difference may be represented as d as shown in FIGS. 2C and 2D. In some embodiments, a front displacement or a rear displacement (such as a range displacement) may exist between image capture devices 122, 124, 126. For example, image capture device 122 may be disposed 0.5 to 2 meters or more behind image capture device 124 and / or image capture device 126. With this type of displacement, one of the image capture devices may be able to cover a potential blind spot of another image capture device. x In some embodiments, a front displacement or a rear displacement (e.g., a range displacement) may exist between image capture devices 122, 124, 126. For example, image capture device 122 may be disposed 0.5 to 2 meters or more behind image capture device 124 and / or image capture device 126. With this type of displacement, one of the image capture devices may be able to cover a potential blind spot of another image capture device.

[0117] The image capture device 122 may have any suitable resolution capability (e.g., the number of pixels associated with the image sensor), and the resolution of the image sensor associated with the image capture device 122 may be higher than, lower than, or the same as the resolution of the image sensors associated with the image capture devices 124 and 126. In some embodiments, the image sensors associated with the image capture device 122 and / or the image capture devices 124 and 126 may have a resolution of 640×480, 1024×768, 1280×960, or any other suitable resolution.

[0118] The frame rate (e.g., the speed at which the image capture device acquires a set of pixel data for one image frame before moving on to the capture of the pixel data associated with the next image frame) may be controllable. The frame rate associated with the image capture device 122 may be higher than, lower than, or the same as the frame rates associated with the image capture devices 124 and 126. The frame rates associated with the image capture devices 122, 124, and 126 may depend on various factors that can affect the timing of the frame rate. For example, one or more of the image capture devices 122, 124, and 126 may include a selectable pixel delay period imposed before or after the acquisition of image data associated with one or more pixels of the image sensor within the image capture device 122, 124, and / or 126. Generally, the image data corresponding to each pixel may be acquired according to the device's clock rate (e.g., 1 pixel per clock cycle). Further, in embodiments including a rolling shutter, one or more of the image capture devices 122, 124, and 126 may include a selectable horizontal blanking period imposed before or after the acquisition of image data associated with the pixel rows of the image sensor within the image capture device 122, 124, and / or 126. Further, one or more of the image capture devices 122, 124, and / or 126 may include a selectable vertical blanking period imposed before or after the acquisition of image data associated with the image frames of the image capture devices 122, 124, and 126.

[0119] With these timing controls, even when the line scan rates of the respective image capture devices are different, it may be possible to synchronize the frame rates associated with the image capture devices 122, 124, and 126. Further, among factors (such as image sensor resolution, maximum line scan rate, etc.), in particular, with these selectable timing controls, even when the field of view of the image capture device 122 is different from the FOVs of the image capture devices 124 and 126, it may be possible to synchronize the image capture from an area where the FOV of the image capture device 122 overlaps with one or more of the FOVs of the image capture devices 124 and 126.

[0120] The frame rate timing at the image capture devices 122, 124, and 126 may depend on the resolution of the associated image sensor. For example, assuming that the line scan rates of both devices are similar, 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, it takes longer to acquire a frame of image data from the sensor with the higher resolution.

[0121] Other factors that can affect the timing of image data acquisition at image capture devices 122, 124, and 126 are the maximum line scan rates. For example, the acquisition of image data lines from the image sensors included in image capture devices 122, 124, and 126 requires some minimum amount of time. Assuming no additional pixel delay period, this minimum amount of time to acquire an image data line will be related to the maximum line scan rate of a particular device. Devices that provide a higher maximum line scan rate have the potential to provide a higher frame rate than devices with a lower maximum line scan rate. In some embodiments, one or both of image capture devices 124 and 126 may have a maximum line scan rate higher than the maximum line scan rate associated with image capture device 122. In some embodiments, the maximum line scan rate of image capture device 124 and / or 126 may be 1.25 times, 1.5 times, 1.75 times, or 2 times or more the maximum line scan rate of image capture device 122.

[0122] In another embodiment, image capture devices 122, 124, and 126 may have the same maximum line scan rate, but image capture device 122 may operate at a scan rate below its maximum scan rate. The system may be configured such that one or both of image capture device 124 and image capture device 126 operate at a line scan rate equal to the line scan rate of image capture device 122. In other examples, the system may be configured such that the line scan rate of image capture device 124 and / or image capture device 126 may be 1.25 times, 1.5 times, 1.75 times, or 2 times or more the line scan rate of image capture device 122.

[0123] In some embodiments, the image capture devices 122, 124, and 126 can be asymmetric. That is, these image capture devices can include cameras with different fields of view (FOV) and focal lengths. The fields of view of the image capture devices 122, 124, and 126 can include, for example, any desired area with respect to the environment of the vehicle 200. In some embodiments, one or more of the image capture devices 122, 124, and 126 can be configured to acquire image data from the environment in front of the vehicle 200, the environment behind the vehicle 200, the environments on both sides of the vehicle 200, or combinations thereof.

[0124] Furthermore, the focal length associated with each image capture device 122, 124, and / or 126 can be selectable (e.g., by inclusion of an appropriate lens, etc.) such that each device can acquire an image of an object within a desired distance range from the vehicle 200. For example, in some embodiments, the image capture devices 122, 124, and 126 can acquire images of nearby objects within a few meters from the vehicle. The image capture devices 122, 124, 126 can also be configured to acquire images of objects in a more distant range from the vehicle (e.g., 25 m, 50 m, 100 m, 150 m, or more). Furthermore, the focal lengths of the image capture devices 122, 124, and 126 can be selected such that one image capture device (e.g., image capture device 122) can acquire an image of an object relatively close to the vehicle (e.g., within 10 m or within 20 m), and other image capture devices (e.g., image capture devices 124 and 126) can acquire images of objects farther from the vehicle 200 (e.g., more than 20 m, more than 50 m, more than 100 m, more than 150 m, etc.).

[0125] According to some embodiments, the FOVs of one or more image capture devices 122, 124, and 126 may have a wide angle. For example, it may be advantageous for the image capture devices 122, 124, and 126, which can be used particularly for capturing images of the area near the vehicle 200, to have a FOV of 140 degrees. For example, the image capture device 122 may be used to capture images of the right or left area of the vehicle 200, and in such embodiments, it may be desirable for the image capture device 122 to have a wide FOV (e.g., at least 140 degrees).

[0126] The field of view associated with each of the image capture devices 122, 124, and 126 may depend on each focal length. For example, as the focal length increases, the corresponding field of view decreases.

[0127] The image capture devices 122, 124, and 126 may be configured to have any suitable field of view. In a specific example, the image capture device 122 may have a horizontal FOV of 46 degrees, the image capture device 124 may have a horizontal FOV of 23 degrees, and the image capture device 126 may have a horizontal FOV of 23 - 46 degrees. In another example, the image capture device 122 may have a horizontal FOV of 52 degrees, the image capture device 124 may have a horizontal FOV of 26 degrees, and the image capture device 126 may have a horizontal FOV of 26 - 52 degrees. In some embodiments, the ratio of the FOV of the image capture device 122 to the FOV of the image capture device 124 and / or the image capture device 126 may vary from 1.5 to 2.0. In other embodiments, this ratio may vary from 1.25 to 2.25.

[0128] System 100 can be configured such that the field of view of image capture device 122 at least partially or completely overlaps with the field of view of image capture device 124 and / or image capture device 126. In some embodiments, System 100 can be configured such that the fields of view of image capture devices 124 and 126 enter, for example, within the field of view of image capture device 122 (e.g., are smaller than the field of view of image capture device 122) and share a common center with the field of view of image capture device 122. In other embodiments, image capture devices 122, 124, and 126 can capture adjacent FOVs or can have partially overlapping FOVs. In some embodiments, the fields of view of image capture devices 122, 124, and 126 can be aligned such that the centers of the narrower FOV image capture devices 124 and / or 126 can be positioned in the lower half of the field of view of the wider FOV device 122.

[0129] FIG. 2F is a graphical representation of an exemplary vehicle control system according to the disclosed embodiments. As shown in FIG. 2F, vehicle 200 can include a throttle system 220, a brake system 230, and a steering system 240. System 100 can provide an input (e.g., a control signal) to one or more of throttle system 220, brake system 230, and steering system 240 via one or more data links (e.g., one or more arbitrary wired links and / or wireless links or links that transmit data). For example, based on the analysis of the images acquired by image capture devices 122, 124, and / or 126, System 100 can provide a control signal for navigating vehicle 200 to one or more of throttle system 220, brake system 230, and steering system 240 (e.g., by causing acceleration, turning, lane shifting, etc.). Further, System 100 can receive an input (e.g., speed, whether vehicle 200 is braking and / or turning, etc.) indicating the operating status of vehicle 200 from one or more of throttle system 220, brake system 230, and steering system 24. Details will be provided below in relation to FIGS. 4-7.

[0130] As shown in FIG. 3A, vehicle 200 can also include a user interface 170 that interacts with the driver or passengers of vehicle 200. For example, the user interface 170 within the vehicle application can include a touch screen 320, a knob 330, buttons 340, and a microphone 350. The driver or passengers of vehicle 200 can also interact with system 100 using a steering wheel (e.g., disposed on or near the steering column of vehicle 200, including, for example, a turn signal steering wheel) and buttons (e.g., disposed on the steering wheel of vehicle 200). In some embodiments, the microphone 350 can be positioned adjacent to the rearview mirror 310. Similarly, in some embodiments, the image capture device 122 can be disposed in the vicinity of the rearview mirror 310. In some embodiments, the user interface 170 can also include one or more speakers 360 (e.g., speakers of the vehicle audio system). For example, system 100 can provide various notifications (e.g., alerts) via the speakers 360.

[0131] FIGS. 3B-3D are diagrams of an exemplary camera mount 370 configured to be positioned behind a rearview mirror (e.g., rearview mirror 310) and facing the vehicle windshield according to the disclosed embodiments. As shown in FIG. 3B, the camera mount 370 can include image capture devices 122, 124, and 126. The image capture devices 124 and 126 can be positioned behind the glare shield 380, which can be in direct contact with the vehicle windshield and can include a composition of film and / or anti-reflective material. For example, the glare shield 380 can be positioned such that the shield is aligned facing the vehicle windshield having a matching slope. In some embodiments, each of the image capture devices 122, 124, and 126 can be positioned behind the glare shield 380, as shown, for example, in FIG. 3D. The disclosed embodiments are not limited to any particular configuration of the image capture devices 122, 124, and 126, the camera mount 370, or the glare shield 380. FIG. 3C is a view of the camera mount 370 shown in FIG. 3B as seen from the front.

[0132] As will be understood by those skilled in the art who benefit from the present disclosure, many variations and / or modifications can be made to the above-described disclosed embodiments. For example, not all components are essential for the operation of system 100. Further, any component can be arranged in any suitable part of system 100, and the components can be rearranged in various configurations while providing the functions of the disclosed embodiments. Accordingly, the configurations discussed above are examples, and regardless of the configurations described above, system 100 can provide a wide range of functions for analyzing the surroundings of vehicle 200 and navigating vehicle 200 in response to the analysis.

[0133] As will be discussed in more detail below, various disclosed embodiments can provide system 100 with various features related to autonomous driving and / or driver assistance technologies. For example, system 100 can analyze image data, position data (e.g., GPS location information), map data, speed data, and / or data from sensors included in vehicle 200. System 100 can collect data for analysis from, for example, image acquisition unit 120, position sensor 130, and other sensors. Further, system 100 can analyze the collected data to determine whether vehicle 200 should take a particular action, and then automatically take the determined action without human intervention. For example, when vehicle 200 is navigating without human intervention, system 100 can automatically control the brakes, acceleration, and / or steering of vehicle 200 (e.g., by sending control signals to one or more of throttle system 220, brake system 230, and steering system 240). Further, system 100 can analyze the collected data and issue warnings and / or alerts to the vehicle occupants based on the analysis of the collected data. Further details regarding the various embodiments provided by system 100 are provided below.

[0134] Forward multi-imaging system

[0135] As discussed above, system 100 may provide a driving assistance function using a multi-camera system. The multi-camera system may use one or more cameras facing the front direction of the vehicle. In other embodiments, the multi-camera system may include one or more cameras facing the side or the rear of the vehicle. In one embodiment, for example, system 100 may use a two-camera imaging system, in which case the first camera and the second camera (e.g., image capture devices 122 and 124) may be positioned at the front and / or side of the vehicle (e.g., vehicle 200). The first camera may have a field of view that is larger than, smaller than, or partially overlapping with the field of view of the second camera. Further, the first camera may be connected to a first image processor to perform monocular image analysis of the image provided by the first camera, and the second camera may be connected to a second image processor to perform monocular image analysis of the image provided by the second camera. The outputs (e.g., processed information) of the first and second image processors may be combined. In some embodiments, the second image processor may receive images from both the first camera and the second camera and perform stereo analysis. In another embodiment, system 100 may use a three-camera imaging system, in which case each camera has a different field of view. Thus, such a system may make a determination based on information derived from objects located at various distances in both the front and side of the vehicle. The reference to monocular image analysis may refer to the case where image analysis is performed based on an image captured from a single viewpoint (e.g., a single camera). Stereo image analysis may refer to the case where image analysis is performed based on two or more images captured with one or more of the image capture parameters changed. For example, the captured images suitable for performing stereo image analysis may include images captured from two or more different positions, images captured from different fields of view, images captured using different focal lengths, images captured with parallax information, and the like.

[0136] For example, in one embodiment, system 100 may implement a three-camera configuration using image capture devices 122, 124, and 126. In such a configuration, image capture device 122 may provide a narrow field of view (e.g., 34 degrees or other values selected from a range of about 20-45 degrees, etc.), image capture device 124 may provide a wide field of view (e.g., 150 degrees or other values selected from a range of about 100 - about 180 degrees), and image capture device 126 may provide a medium field of view (e.g., 46 degrees or other values selected from a range of about 35 - about 60 degrees). In some embodiments, image capture device 126 may operate as the main or primary camera. Image capture devices 122, 124, and 126 may be positioned substantially side by side (e.g., separated by 6 cm) behind rearview mirror 310. Further, in some embodiments, as discussed above, one or more of image capture devices 122, 124, and 126 may be mounted behind glare shield 380 that is in the same plane as the front windshield of vehicle 200. Such a shield may operate to minimize any reflections from inside the vehicle onto image capture devices 122, 124, and 126.

[0137] In another embodiment, as discussed above in connection with FIGS. 3B and 3C, the wide field of view camera (e.g., image capture device 124 in the above example) may be mounted lower than the narrow main field of view cameras (e.g., image capture devices 122 and 126 in the above example). This configuration may provide a clear line of sight from the wide field of view camera. To reduce reflections, the camera may be mounted near the front windshield of vehicle 200 and may include a polarizer in the camera to attenuate the reflected light.

[0138] The three-camera system may provide certain performance features. For example, some embodiments may include the ability to verify the detection of an object by one camera based on the detection results from another camera. In the three-camera configuration discussed above, processing unit 110 may include, for example, three processing devices (e.g., three EyeQ series processor chips as discussed above), and each processing device may be directed to process the images captured by one or more of image capture devices 122, 124, and 126.

[0139] In a three-camera system, a first processing device may receive images from both the main camera and the narrow field of view camera, perform vision processing on the narrow FOV camera, and detect, for example, other vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. Further, the first processing device may calculate the pixel discrepancies between the image from the main camera and the image from the narrow camera and create a 3D reconstruction of the environment of vehicle 200. Next, the first processing device may combine the 3D reconstruction with 3D information calculated based on 3D map data or information from another camera.

[0140] A second processing device may receive an image from the main camera, perform vision processing, and detect other vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. Further, the second processing device may calculate the camera displacement and, based on the displacement, calculate the pixel discrepancies between consecutive images and create a 3D reconstruction of the scene (e.g., structure from motion). The second processing device may send the structure from motion based on the 3D reconstruction to the first processing device and combine the structure from motion with a stereoscopic 3D image.

[0141] A third processing device may receive an image from the wide FOV camera, process the image, and detect vehicles, pedestrians, lane markings, traffic signs, traffic lights, and other road objects. The third processing device may further execute additional processing instructions to analyze the image and identify moving objects within the image, such as vehicles and pedestrians during a lane change.

[0142] In some embodiments, independently capturing and processing an image-based information stream may provide an opportunity for redundancy in the system. Such redundancy may include, for example, verifying and / or capturing information obtained by capturing and processing image information from at least a second image capture device using a first image capture device and the images processed from that device.

[0143] In some embodiments, system 100 may use two image capture devices (e.g., image capture devices 122 and 124) to provide navigation assistance to vehicle 200, and use a third image capture device (e.g., image capture device 126) to provide redundancy and verify the analysis of data received from the other two image capture devices. For example, in such a configuration, image capture devices 122 and 124 may provide images for stereoscopic analysis by system 100 to navigate vehicle 200, and image capture device 126 may provide images for monocular analysis by system 100 to provide redundancy and verification of information obtained based on images captured from image capture device 122 and / or image capture device 124. That is, image capture device 126 (and corresponding processing device) may be considered to provide a redundancy subsystem that provides a check on the analysis derived from image capture devices 122 and 124 (e.g., to provide an automatic emergency braking (AEB) system). Further, in some embodiments, the redundancy and verification of received data can 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, etc.).

[0144] Those skilled in the art will recognize that the above camera configurations, camera arrangements, number of cameras, camera positions, etc. are merely illustrative. These components and the like described for the overall system can be assembled and used in various 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 follow below.

[0145] FIG. 4 is an exemplary functional block diagram of memory 140 and / or 150 that may store / program instructions to execute one or more operations according to the disclosed embodiments. Although memory 140 is referred to below, those skilled in the art will recognize that the instructions may be stored in memory 140 and / or 150.

[0146] As shown in FIG. 4, the memory 140 may store a monocular image analysis module 402, a stereo image analysis module 404, a speed and acceleration module 406, and a navigation response module 408. The disclosed embodiments are not limited to any particular configuration of the memory 140. Further, the application processor 180 and / or the image processor 190 may execute instructions stored in any of the modules 402, 404, 406, and 408 included in the memory 140. Those skilled in the art will understand that references to the processing unit 110 in the following discussion may refer to the application processor 180 and the image processor 190 individually or together. Accordingly, any of the steps of the following processes may be executed by one or more processing devices.

[0147] In one embodiment, the monocular image analysis module 402 may store instructions (such as computer vision software) for performing monocular image analysis of a set of images acquired by one of the image capture devices 122, 124, and 126 when executed by the processing unit 110. In some embodiments, the processing unit 110 may combine information from the set of images with additional sensory information (such as information from radar, lidar, etc.) to perform monocular image analysis. As will be described below in connection with FIGS. 5A-5D, the monocular image analysis module 402 may include instructions for detecting a set of features within the set of images, such as lane marks, vehicles, pedestrians, road signs, highway exit ramps, traffic signals, hazards, and any other features associated with the vehicle's environment. Based on the analysis, the system 100 may cause one or more navigation responses, such as turns, lane shifts, and acceleration changes, to occur in the vehicle 200 (e.g., via the processing unit 110), as will be discussed below in connection with the navigation response module 408.

[0148] In one embodiment, the stereo image analysis module 404 may store instructions (such as computer vision software), which, when executed by the processing unit 110, perform stereo image analysis of a first and a second set of images obtained by a combination of image capture devices selected from the image capture devices 122, 124, and 126. In some embodiments, the processing unit 110 may combine information from the first and second sets of images with additional sensory information (e.g., information from radar) to perform stereo image analysis. For example, the stereo image analysis module 404 may include instructions to perform stereo image analysis based on a first set of images obtained by the image capture device 124 and a second set of images obtained by the image capture device 126. As will be described below in connection with FIG. 6, the stereo image analysis module 404 may include instructions to detect sets of features in the first and second sets of images such as lane markings, vehicles, pedestrians, road signs, highway exit ramps, traffic signals, and hazards. Based on the analysis, the processing unit 110 may cause one or more navigation responses, such as turns, lane shifts, and acceleration changes, to occur in the vehicle 200, as will be described later in connection with the navigation response module 408. Further, in some embodiments, the stereo image analysis module 404 may implement techniques related to trained systems (such as neural networks or deep neural networks) or untrained systems such as systems configured to detect and / or label objects in an environment in which sensory information is captured and processed using computer vision algorithms. 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.

[0149] In one embodiment, the speed and acceleration module 406 may store software configured to analyze data received from one or more computing and electromechanical devices within the vehicle 200 configured to alter the speed and / or acceleration of the vehicle 200. For example, the processing unit 110 may execute instructions associated with the speed and acceleration module 406 to calculate a target speed of 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, speed and / or acceleration, nearby vehicles, pedestrians or road objects, the position and / or speed of the vehicle 200 relative thereto, and position information of the vehicle 200 relative to the lane markings of the road. Additionally, the processing unit 110 may calculate the target speed of the vehicle 200 based on sensory input (e.g., information from radar) and inputs from other systems of the vehicle 200, such as the throttle system 220, the brake system 230, and / or the steering system 240 of the vehicle 200. Based on the calculated target speed, the processing unit 110 may send an electronic signal to the throttle system 220, the brake system 230, and / or the steering system 240 of the vehicle 200 to trigger a change in speed and / or acceleration, for example, by physically weakening the brakes of the vehicle 200 or by weakening the accelerator.

[0150] In one embodiment, the navigation response module 408 may store executable software by the processing unit 110 and 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, as well as target position information of the vehicle 200 and the like. Further, in some embodiments, the navigation response may be (partially or fully) based on map data, a predetermined position of the vehicle 200, and / or 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 sensory input (e.g., information from radar) and input from other systems of the vehicle 200 such as the throttle system 220, the brake system 230, and the steering system 240 of the vehicle 200. Based on the desired navigation response, the processing unit 110 may transmit an electronic signal to the throttle system 220, the brake system 230, and the steering system 240 of the vehicle 200 to trigger the desired navigation response, for example, by turning the steering wheel of the vehicle 200 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 an input to the execution of the speed and acceleration module 406 for calculating a speed change of the vehicle 200.

[0151] Furthermore, any of the modules disclosed herein (e.g., modules 402, 404, and 406) may implement techniques related to a trained system (such as a neural network or a deep neural network) or an untrained system.

[0152] FIG. 5A is a flowchart showing an exemplary process 500A that generates one or more navigation responses based on monocular image analysis according to the disclosed embodiments. At step 510, the processing unit 110 may receive a plurality of 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 (such as the image capture device 122 having a field of view 202) may capture a plurality of images of an area in front of (or for example, the side or rear of) the vehicle 200 and transmit them to the processing unit 110 via a data connection (such as digital, wired, USB, wireless, Bluetooth, etc.). The processing unit 110 may execute the monocular image analysis module 402 to analyze the plurality of images at step 520, as will be described in more detail below with reference to FIGS. 5B-5D. By performing the analysis, the processing unit 110 may detect a set of features within a set of images of lane marks, vehicles, pedestrians, road signs, highway exit ramps, traffic signals, etc.

[0153] At step 520, the processing unit 110 may also execute the monocular image analysis module 402 to detect various road hazards such as parts of truck tires, fallen road signs, loose cargo, and small animals. The structure, shape, size, and color of road hazards can vary, making the detection of such hazards more difficult. In some embodiments, the processing unit 110 may execute the monocular image analysis module 402 to perform multi-frame analysis on the plurality of images to detect road hazards. For example, the processing unit 110 may estimate the movement of the camera between consecutive image frames, calculate the pixel discrepancies between the frames, and construct a 3D map of the road. Next, the processing unit 110 may use the 3D map to detect the road surface and hazards present on the road surface.

[0154] In step 530, the processing unit 110 executes the navigation response module 408 to cause one or more navigation responses in the vehicle 200 based on the analysis performed in step 520 and the techniques described above in connection with FIG. 4. The navigation responses can include, for example, turns, lane shifts, and acceleration changes. In some embodiments, the processing unit 110 can use data derived from the execution of the speed and acceleration module 406 to cause one or more navigation responses. Further, the multiple navigation responses can be performed simultaneously, sequentially, or in any combination thereof. For example, the processing unit 110 can cause the vehicle 200 to cross one lane and then, for example, accelerate by sequentially transmitting control signals to the steering system 240 and the throttle system 220 of the vehicle 200. Alternatively, the processing unit 110 can cause the vehicle 200 to apply the brakes and simultaneously shift lanes by, for example, simultaneously transmitting control signals to the brake system 230 and the steering system 240 of the vehicle 200.

[0155] FIG. 5B is a flowchart illustrating an exemplary process 500B for detecting one or more vehicles and / or pedestrians within a set of images according to the disclosed embodiments. The processing unit 110 can execute the monocular image analysis module 402 to implement the process 500B. In step 540, the processing unit 110 can identify a set of candidate objects representing vehicles and / or pedestrians that may be present. For example, the processing unit 110 can scan one or more images, compare the images to one or more predetermined patterns, and identify positions within each image that may contain target objects (e.g., vehicles, pedestrians, or portions thereof). The predetermined patterns can be specified to achieve a low rate of "false positives" and a low rate of "misses". For example, the processing unit 110 can use a low similarity threshold to a predetermined pattern to identify candidate objects as potential vehicles or pedestrians. By doing so, the processing unit 110 can potentially reduce the probability of missing (e.g., not identifying) candidate objects representing vehicles or pedestrians.

[0156] In step 542, the processing unit 110 may filter the set of candidate objects to exclude certain candidates (e.g., irrelevant or low - relevance objects) based on classification criteria. Such criteria can be derived from various characteristics associated with the object types stored in a database (e.g., the database stored in the memory 140). The characteristics may include the shape, dimensions, texture, and position of the object (e.g., relative to the vehicle 200), etc. Thus, the processing unit 110 may use one or more sets of criteria to reject false candidates from the set of candidate objects.

[0157] In step 544, the processing unit 110 may analyze a plurality of image frames to determine whether the objects within the set of candidate objects represent a vehicle and / or a pedestrian. For example, the processing unit 110 may track the candidate objects detected over consecutive frames and accumulate frame - by - frame data (e.g., size, position relative to the vehicle 200, etc.) associated with the detected objects. Further, the processing unit 110 may estimate the parameters of the detected objects and compare the frame - by - frame position data of the objects with the predicted positions.

[0158] In step 546, the processing unit 110 may construct a set of measurements of the detected object. Such measurements may include, for example, position, velocity, and acceleration values (with respect to the vehicle 200) associated with the detected object. In some embodiments, the processing unit 110 is based on an estimation technique that uses a series of time-based observations such as a Kalman filter or linear quadratic estimation (LQE) and / or based on modeling data available for different object types (e.g., cars, trucks, pedestrians, bicycles, road signs, etc.). The Kalman filter may be based on a measurement of the scale of the object, where the scale measurement is proportional to the time to collision (e.g., the amount of time until the vehicle 200 reaches the object). Thus, by performing steps 540-546, the processing unit 110 may identify vehicles and pedestrians appearing within the set of captured images and derive information (e.g., position, velocity, size) associated with the vehicles and pedestrians. Based on the identified and derived information, the processing unit 110 may cause one or more navigation responses in the vehicle 200 as described above in connection with FIG. 5A.

[0159] In step 548, the processing unit 110 may perform an optical flow analysis of 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 a movement pattern different from the movement of the road surface with respect to the vehicle 200 within one or more images associated with other vehicles and pedestrians. The processing unit 110 may calculate the movement of the candidate object by observing different positions of the object over multiple image frames captured at different times. The processing unit 110 may calculate the movement of the candidate object using the position and time values as inputs to a mathematical model. Thus, optical flow analysis may provide another way to detect vehicles and pedestrians in the vicinity of the vehicle 200. The processing unit 110 may perform optical flow analysis in combination with steps 540-546 to provide redundancy in detecting vehicles and pedestrians and increase the reliability of the system 100.

[0160] FIG. 5C is a flowchart illustrating an exemplary process 500C for detecting road markings and / or lane geometry information within a set of images according to the disclosed embodiments. The processing unit 110 may execute the monocular image analysis module 402 to implement the process 500C. At step 550, the processing unit 110 may detect a set of objects by scanning one or more images. To detect lane mark segments, lane geometry information, and other related road markings, the processing unit 110 may filter the set of objects to exclude those determined to be irrelevant (e.g., small holes, small rocks, etc.). At step 552, the processing unit 110 may group together the segments detected at step 550 that belong to the same road marking or lane mark. Based on the grouping, the processing unit 110 may develop a model, such as a mathematical model, to represent the detected segments.

[0161] At step 554, the processing unit 110 may construct a set of measurements associated with the detected segments. In some embodiments, the processing unit 110 may create a projection of the detected segments from the image plane to the real-world plane. The projection may be characterized using a cubic polynomial having coefficients corresponding to physical characteristics such as the position, slope, curvature, and curvature derivative of the detected road. In generating the projection, the processing unit 110 may take into account road surface variations as well as the pitch and roll rates associated with the vehicle 200. Additionally, the processing unit 110 may model the road height by analyzing the position and motion cues present on the road surface. Further, the processing unit 110 may estimate the pitch rate and roll rate associated with the vehicle 200 by tracking a set of feature points in one or more images.

[0162] In step 556, the processing unit 110 may perform multi-frame analysis, for example, by tracking the detection section over consecutive image frames and accumulating frame-by-frame data associated with the detection section. When the processing unit 110 performs multi-frame analysis, the set of measurements constructed in step 554 may become more reliable and may be associated with an increasingly high confidence level. Therefore, by executing steps 550, 552, 554, and 556, the processing unit 110 may identify road markings appearing within the set of captured images and derive lane geometry information. Based on the identified and derived information, the processing unit 110 may cause one or more navigation responses in the vehicle 200 as described above in connection with FIG. 5A.

[0163] In step 558, the processing unit 110 may further develop a safety model of the vehicle 200 in the context of the situation around the vehicle, taking into account additional information sources. The processing unit 110 may use the safety model to define situations in which the system 100 can safely perform autonomous control of the vehicle 200. To develop the safety model, in some embodiments, the processing unit 110 may consider the position and movement of other vehicles, the detected road edges and barriers, and / or a general road shape description extracted from map data (such as data from the map database 160). By considering additional information sources, the processing unit 110 may provide redundancy in detecting road markings and lane geometry and increase the reliability of the system 100.

[0164] FIG. 5D is a flowchart showing an exemplary process 500D for detecting traffic lights within a set of images according to the disclosed embodiment. The processing unit 110 may execute the monocular image analysis module 402 to implement the process 500D. At step 560, the processing unit 110 may scan a set of images and identify objects that appear at positions within the images that are likely to include traffic lights. For example, the processing unit 110 may filter the identified objects to construct a set of candidate objects excluding objects that are unlikely to correspond to traffic lights. The filtering may be performed based on various characteristics associated with traffic lights, such as shape, dimension, texture, and position (e.g., relative to the vehicle 200). Such characteristics may be based on many examples of traffic lights and traffic control signals and may be stored in a database. In some embodiments, the processing unit 110 may perform multi-frame analysis on the set of candidate objects reflecting possible traffic lights. For example, the processing unit 110 may track candidate objects over consecutive image frames, estimate the real-world positions of the candidate objects, and filter out (objects that are unlikely to be traffic lights) moving objects. In some embodiments, the processing unit 110 may perform color analysis on the candidate objects and identify the relative positions of the detected colors represented within the possible traffic lights.

[0165] At step 562, the processing unit 110 may analyze the geometry of the intersection. The analysis may be based on any combination of (i) the number of lanes detected on both sides of the vehicle 200, (ii) the marks (such as arrow marks) detected on the road, and (iii) the description of the intersection extracted from map data (such as data from the map database 160). The processing unit 110 may perform the analysis using the information derived from the execution of the monocular analysis module 402. In addition, the processing unit 110 may identify the correspondence between the traffic lights detected at step 560 and the lanes that appear near the vehicle 200.

[0166] As the vehicle 200 approaches an intersection, at step 564, the processing unit 110 may update the intersection geometry being analyzed and the reliability associated with the detected traffic signal. For example, the number of traffic signals estimated to appear at the intersection compared to the actual number of traffic signals that appear at the intersection may affect the reliability. Thus, based on the reliability, the processing unit 110 may delegate control to the driver of the vehicle 200 to improve the safety situation. By executing steps 560, 562, and 564, the processing unit 110 may identify the traffic signals that appear within the set of captured images and analyze the intersection geometry information. Based on the identification and analysis, the processing unit 110 may cause one or more navigation responses to occur in the vehicle 200 as described above in connection with FIG. 5A.

[0167] FIG. 5E is a flowchart of an exemplary process 500E for causing one or more navigation responses to occur in the vehicle 200 based on a vehicle route, according to the disclosed embodiment. At step 570, the processing unit 110 may construct an initial vehicle route associated with the vehicle 200. The vehicle route may be represented using a set of points represented by coordinates (x, y), and the distance d between two points within the set of points i may be in the range of 1 to 5 meters. In one embodiment, the processing unit 110 may construct the initial vehicle route using two polynomials such as left and right road polynomials. The processing unit 110 may calculate the geometric midpoint between the two polynomials and, if there is a predetermined offset (offset 0 may correspond to driving in the center of the lane), offset each point included in the resulting vehicle route by a predetermined offset (e.g., a smart lane offset). The offset may be in a direction perpendicular to the section between any two points within the vehicle route. In another embodiment, the processing unit 110 may use one polynomial and an estimated lane width to offset each point of the vehicle route by only the sum of a predetermined offset (e.g., a smart lane offset) and half of the estimated lane width.

[0168] 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 within the set of points representing the vehicle route k so that it is shorter than the above-mentioned distance d i and may reconstruct the vehicle route constructed in step 570 using a higher resolution. For example, the distance d k may be in the range of 0.1 to 0.3 meters. The processing unit 110 may reconstruct the vehicle route using a parabolic spline algorithm, which may result in a cumulative distance vector S corresponding to the total length of the vehicle route (i.e., based on the set of points representing the vehicle route).

[0169] In step 574, the processing unit 110 may identify a look-ahead point (represented in coordinates as (x l , z l )) based on the updated vehicle route constructed in step 572. The processing unit 110 may extract the look-ahead point from the cumulative distance vector S and may associate a look-ahead distance and a look-ahead time with the look-ahead point. The look-ahead distance may have a lower limit range of 10 to 20 meters and may be calculated as the product of the speed of the vehicle 200 and the look-ahead time. For example, as the speed of the vehicle 200 decreases, the look-ahead distance may also decrease (e.g., until it reaches the lower limit). The look-ahead time, which may be in the range of 0.5 to 1.5 seconds, may be inversely proportional to the gain of one or more control loops associated with generating a navigation response in the vehicle 200, such as a progress error tracking control loop. For example, the gain of the progress error tracking control loop may depend on the bandwidths of a yaw rate loop, a steering actuator loop, and vehicle lateral dynamics. Therefore, the higher the gain of the progress error tracking control loop, the shorter the look-ahead time.

[0170] In step 576, the processing unit 110 may determine a progress error and a yaw rate command based on the look-ahead point identified in step 574. The processing unit 110 calculates the arctangent of the look-ahead point, e.g., arctan(x l / z lBy calculating [[ID=]], the progress error can be identified. The processing unit 110 may determine a yaw rate command as the product of the progress error and the high-level control gain. The high-level control gain may be a value equal to (2 / look-ahead time) when the look-ahead distance is not at the lower limit. When the look-ahead distance is at the lower limit, the high-level control gain may be a value equal to (2 * the speed of the vehicle 200 / look-ahead distance).

[0171] FIG. 5F is a flowchart showing an exemplary process 500F for identifying whether a preceding vehicle is changing lanes according to the disclosed embodiment. In step 580, the processing unit 110 may identify navigation information associated with a preceding vehicle (e.g., a vehicle traveling in front of vehicle 200). For example, the processing unit 110 may identify the position, speed (e.g., direction and speed) and / or acceleration of the preceding vehicle using the techniques described above in connection with FIGS. 5A and 5B. The processing unit 110 may also identify one or more road polynomials, look-ahead points (associated with vehicle 200) and / or snail trails (e.g., a set of points describing the path taken by the preceding vehicle) using the techniques described above in connection with FIG. 5E.

[0172] In step 582, the processing unit 110 may analyze the navigation information identified in step 580. In one embodiment, the processing unit 110 may calculate the distance (e.g., along the trail) between the snail trail and the road polynomial. If the difference in this distance along the trail exceeds a predetermined threshold (e.g., 0.1 - 0.2 meters for a straight road, 0.3 - 0.4 meters for a gently curved road, 0.5 - 0.6 meters for a sharp - curved road), the processing unit 110 may determine that the leading vehicle is likely to be changing lanes. When it is detected that a plurality of vehicles are traveling in front 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 be changing lanes. The processing unit 110 may further compare the curvature of the snail trail (associated with the leading vehicle) with the expected curvature of the road segment on which the leading vehicle is traveling. The expected curvature can be extracted from map data (e.g., data from the map database 160), the road polynomial, the snail trails of other vehicles, and prior knowledge about the road, etc. 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 leading vehicle is likely to be changing lanes.

[0173] In another embodiment, the processing unit 110 may compare the instantaneous position of the leading vehicle with the look - ahead point (associated with vehicle 200) over a specific time period (e.g., 0.5 - 1.5 seconds). If the sum of the differences in distance and divergence between the instantaneous position of the leading vehicle and the look - ahead point during the specific time period exceeds a predetermined threshold (e.g., 0.3 - 0.4 meters for a straight road, 0.7 - 0.8 meters for a gently curved road, 1.3 - 1.7 meters for a sharp - curved road), the processing unit 110 may determine that the leading vehicle is likely to be changing lanes. In another embodiment, the processing unit 110 may analyze the geometry of the snail trail by comparing the lateral distance traveled along the trail with the expected curvature of the snail trail. The expected radius of curvature is calculated: (δz 2 +δ x 2 ) / 2 / (δ x ) can be specified according to, where σ x represents the lateral movement distance, and σ z represents the longitudinal movement distance. When the difference between the lateral movement distance and the expected curvature exceeds a predetermined threshold (for example, 500 to 700 meters), the processing unit 110 may determine that the preceding vehicle is likely to be changing lanes. In another embodiment, the processing unit 110 may analyze the position of the preceding vehicle. When the position of the preceding vehicle obscures the road polynomial (for example, the preceding vehicle overlaps the road polynomial), the processing unit 110 may determine that the preceding vehicle is likely to be changing lanes. When the position of the preceding vehicle is such that another vehicle is detected in front of the preceding vehicle 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 be changing lanes.

[0174] In step 584, the processing unit 110 may identify whether the preceding vehicle 200 is changing lanes based on the analysis performed in step 582. For example, the processing unit 110 may make its determination based on the weighted average of the individual analyses performed in step 582. Under such a scheme, for example, a determination by the processing unit 110 that the preceding vehicle is likely to be changing 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 be changing lanes). Different weights may be assigned to the different analyses performed in step 582, and the disclosed embodiments are not limited to any particular combination of analyses and weights.

[0175] FIG. 6 is a flowchart showing an exemplary process 600 that generates one or more navigation responses based on stereoscopic image analysis according to the disclosed embodiments. At step 610, the processing unit 110 may receive a first and a second plurality of images via the data interface 128. For example, cameras included in the image acquisition unit 120 (such as the image capture devices 122 and 124 having the fields of view 202 and 204) may capture a first and a second plurality of 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 the first and the second plurality of images via more than two data interfaces. The disclosed embodiments are not limited to any particular data interface configuration or protocol.

[0176] In step 620, the processing unit 110 executes the stereo image analysis module 404 to perform stereo image analysis of the first and second plurality of images, create a 3D map of the road in front of the vehicle, and detect features in the images such as lane marks, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, and road hazards. The stereo image analysis can be performed in the same manner as the steps described above in connection with FIGS. 5A-5D. For example, the processing unit 110 executes the stereo image analysis module 404 to detect candidate objects (e.g., vehicles, pedestrians, road marks, traffic lights, road hazards, etc.) in the first and second plurality of images, filter and exclude a subset of the candidate objects based on various criteria, perform multi-frame analysis, construct measurements, and identify the reliability of the remaining candidate objects. In performing the above steps, the processing unit 110 can consider information from both the first and second plurality of images rather than just information from one set of the images. For example, the processing unit 110 can analyze the difference in pixel-level data (or other data subsets from the two streams of captured images) of candidate objects that appear in both the first and second plurality of images. As another example, the processing unit 110 can estimate the position and / or velocity (e.g., with respect to vehicle 200) of a candidate object by observing that the object appears in one of the plurality of images but not in others, or with respect to other differences that may exist for an object that appears in the case of the two image streams. For example, the position, velocity, and / or acceleration with respect to vehicle 200 can be identified based on the trajectory, position, movement characteristics, etc. of features associated with an object that appears in one or both of the image streams.

[0177] In step 630, the processing unit 110 executes the navigation response module 408 to cause one or more navigation responses in the vehicle 200 based on the analysis executed in step 620 and the techniques described above in connection with FIG. 4. The navigation responses can include, for example, turns, lane shifts, acceleration changes, speed changes, and braking. In some embodiments, the processing unit 110 can use data derived from the execution of the speed and acceleration module 406 to cause one or more navigation responses. Further, the plurality of navigation responses can be performed simultaneously, sequentially, or in any combination thereof.

[0178] FIG. 7 is a flowchart showing an exemplary process 700 for causing one or more navigation responses based on the analysis of three sets of images according to the disclosed embodiments. In step 710, the processing unit 110 can receive a first, a second, and a third plurality of images via the data interface 128. For example, cameras included in the image acquisition unit 120 (such as the image capture devices 122, 124, and 126 having the fields of view 202, 204, and 206) can capture a first, a second, and a third plurality of images of the area in front of and / or to the sides of the vehicle 200 and transmit them to the processing unit 110 via a digital connection (such as USB, wireless, Bluetooth, etc.). In some embodiments, the processing unit 110 can receive the first, the second, and the third plurality of images via more than three data interfaces. For example, each of the image capture devices 122, 124, and 126 can have an associated data interface for communicating data to the processing unit 110. The disclosed embodiments are not limited to any particular data interface configuration or protocol.

[0179] 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 marks, vehicles, pedestrians, road signs, highway exit ramps, traffic lights, and road hazards. The analysis may be performed in the same manner as the steps described above in connection with FIGS. 5A-5D and 6. For example, the processing unit 110 may perform monocular image analysis on each of the first, second, and third plurality of images (e.g., based on the execution of the monocular image analysis module 402 and the steps described above in connection with FIGS. 5A-5D). 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 (e.g., via the execution of the stereo image analysis module 404 and based on the steps described above in connection with FIG. 6). 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 image analysis and stereo image analysis. For example, the processing unit 110 may perform monocular image analysis on the first plurality of images (e.g., via the execution of the monocular image analysis module 402) and perform stereo image analysis on the second and third plurality of images (e.g., via the execution of the stereo image analysis module 404). The configuration of the image capture devices 122, 124, and 126 - including each position and field of view 202, 204, and 206 - may affect the type of analysis performed on the first, second, and third plurality of images. The disclosed embodiments are not limited to a particular configuration of the image capture devices 122, 124, and 126 or the type of analysis performed on the first, second, and third plurality of images.

[0180] 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 with a particular configuration of the image capture devices 122, 124, and 126. For example, the processing unit 110 may identify the rate of "false positives" (e.g., when the system 100 incorrectly determines the presence of a vehicle or pedestrian) and "misses".

[0181] In step 730, the processing unit 110 may generate one or more navigation responses in the vehicle 200 based on information derived from two of the first, second, and third plurality of images. The selection of two of the first, second, and third plurality of images may depend on various factors such as the number, type, and size of objects detected in each of the plurality of images. The processing unit 110 may make the selection based on, for example, the quality and resolution of the images, the effective field of view reflected in the images, the number of captured frames, and the degree to which one or more objects of interest actually appear in the frames (e.g., the percentage of frames in which the object appears, the proportion in which the object appears in each such frame, etc.).

[0182] In some embodiments, the processing unit 110 may select information derived from two of the first, second, and third plurality of images by determining the degree to which information derived from one image source is consistent with information derived from other image sources. For example, the processing unit 110 may combine the processed information (regardless of whether it is monocular analysis, stereoscopic analysis, or any combination of the two) derived from each of the image capture devices 122, 124, and 126 to identify visual indicators (e.g., lane marks, detected vehicles and / or their positions and / or routes, detected traffic lights, etc.) that are consistent across the images captured from each of the image capture devices 122, 124, and 126. The processing unit 110 may also exclude information that is not consistent across the captured images (e.g., a vehicle changing lanes, a lane model indicating a vehicle too close to the vehicle 200, etc.). Thus, the processing unit 110 may select information derived from two of the first, second, and third plurality of images based on the identification of consistent and inconsistent information.

[0183] Navigation responses can include, for example, turns, lane shifts, and acceleration changes. The processing unit 110 can generate one or more navigation responses based on the analysis executed in step 720 and the techniques described above in connection with FIG. 4. The processing unit 110 can also use data derived from the execution of the speed and acceleration module 406 to generate one or more navigation responses. In some embodiments, the processing unit 110 can generate one or more navigation responses based on the relative position, relative speed, and / or relative acceleration between the vehicle 200 and an object detected in any of the first, second, and third plurality of images. The plurality of navigation responses can be performed simultaneously, sequentially, or in any combination thereof.

[0184] Sparse Road Model for Autonomous Vehicle Navigation

[0185] In some embodiments, the disclosed systems and methods can use a sparse map for autonomous vehicle navigation. Specifically, the sparse map can be for autonomous vehicle navigation along a road segment. For example, the sparse map can provide sufficient information for navigating an autonomous vehicle without storing and / or updating large amounts of data. As discussed in more detail below, an autonomous vehicle can use the sparse map to navigate one or more roads based on one or more stored trajectories.

[0186] Sparse Map for Autonomous Vehicle Navigation

[0187] In some embodiments, the disclosed systems and methods may generate a sparse map for autonomous vehicle navigation. For example, a sparse map may provide sufficient information for navigation without requiring excessive data storage or data transfer speed. As discussed in more detail below, a vehicle (which may be an autonomous vehicle) may use the sparse map to navigate one or more roads. For example, in some embodiments, a sparse map may include data related to a road and potential landmarks along the road that may be sufficient for vehicle navigation but also exhibit a small data footprint. For example, a sparse data map, as described in detail below, may require significantly less storage area and data transfer bandwidth compared to a digital map that includes detailed map information such as image data collected along a road.

[0188] For example, instead of storing a detailed representation of a road segment, a sparse data map may store a three-dimensional polynomial representation of a preferred vehicle route along the road. These routes may require little data storage area. Further, in the sparse data map being described, landmarks may be identified and included in the sparse map road model to assist in navigation. These landmarks may be placed at any interval suitable for enabling vehicle navigation, but in some cases, it may not be necessary to identify and include such landmarks at high density and short intervals. Rather, in some cases, navigation may be possible based on landmarks that are at least 50 meters, at least 100 meters, at least 500 meters, at least 1 kilometer, or at least 2 kilometers apart. As 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 an image capture device, a global positioning system sensor, a motion sensor, etc. as the vehicle moves along a road. In some cases, a sparse map may be generated based on data collected during multiple runs of one or more vehicles along a particular road. Generating a sparse map using multiple runs of one or more vehicles may be referred to as "crowdsourcing" of the sparse map.

[0189] According to the disclosed embodiments, an autonomous vehicle system may use a sparse map for navigation. For example, the disclosed systems and methods may distribute a sparse map for generating a road navigation model for an autonomous vehicle, and use the sparse map and / or the generated road navigation model to navigate the autonomous vehicle along a road segment. The sparse map according to the present disclosure may include one or more three-dimensional contours that may represent a predetermined trajectory that the autonomous vehicle may cross when moving along an associated road segment.

[0190] The sparse map according to the present disclosure may also 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 vehicle navigation. The sparse map according to the present disclosure may enable autonomous navigation of a vehicle based on a relatively small amount of data included in the sparse map. For example, without including a detailed representation of the road, such as data showing in detail the road end, the curvature of the road, an image associated with the road segment, or other physical features associated with the road segment, the disclosed embodiments of the sparse map may require a relatively small memory area (and a relatively small bandwidth when a portion of the sparse map is transferred to the vehicle), but still be able to appropriately provide autonomous vehicle navigation. The small data footprint of the disclosed sparse map, discussed in more detail below, may be achieved in some embodiments by storing a representation of road-related elements that require a small amount of data but still enable autonomous navigation.

[0191] 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 that a vehicle can follow along the road. Thus, rather than storing (or having to transmit) details regarding the physical properties of the road to enable navigation along the road using the disclosed sparse map, a vehicle may instead align its driving route to a trajectory (e.g., a polynomial spline) along a particular road segment without the need to interpret the physical aspects of the road in some cases, and thereby be navigated along a particular road segment. In this way, a vehicle may be navigated based on stored trajectories (e.g., polynomial splines) that may require much less memory than techniques that primarily involve storing road images, road parameters, road layouts, etc.

[0192] In addition to the stored polynomial representations of trajectories along road segments, the disclosed sparse map may also include small data objects that can represent features of the road. In some embodiments, the small data objects may include digital signatures derived from digital images (or digital signals) acquired by sensors (e.g., cameras or other sensors such as suspension sensors) mounted on a vehicle traveling along a road segment. The digital signature may be of a reduced size compared to the signal acquired by the sensor. In some embodiments, the digital signature may be created to be compatible with, for example, a classifier function configured to detect and identify road features from signals acquired by the sensor during its travel. In some embodiments, the digital signature may be created to have the smallest possible footprint while maintaining the ability to correlate or match road features to the stored signature based on images of road features captured by a camera mounted on a vehicle traveling along the same road segment at a later time (or, if the stored signature is not image-based and / or includes other data, digital signals generated by the sensor).

[0193] In some embodiments, the size of the data object may be further associated with the uniqueness of the road feature. For example, for road features detectable by a camera mounted on a vehicle, if the vehicle-mounted camera system is coupled to a classifier that can distinguish the image data corresponding to such road features as being associated with a particular type of road feature, e.g., a road sign, and such road signs are locally unique in that area (e.g., there are no identical road signs or road signs of the same type in the vicinity), it may be sufficient to store only the data indicating the type and location of the road feature.

[0194] As discussed in more detail below, road features (e.g., landmarks along a road segment) can be stored as small data objects that can represent the road feature in relatively few bytes, while at the same time providing sufficient information for recognizing and using such features for navigation. In one example, a road sign can be identified as a recognized landmark on which vehicle navigation can be based. The representation of the road sign can be stored in a sparse map such that it includes, for example, a few bytes of data indicating the type of the landmark (e.g., a stop sign) and a few bytes of data indicating the location of the landmark (e.g., coordinates). Navigating based on such a data perspective representation of the landmark (e.g., using a representation sufficient to identify, recognize, and navigate based on the landmark) can provide the desired level of navigation functionality associated with the sparse map without significantly increasing the data overhead associated with the sparse map. Such an efficient representation of such landmarks (and other road features) can utilize such vehicle-mounted sensors and processors configured to detect, identify, and / or classify particular road features.

[0195] For example, when a landmark or a particular type of landmark is locally unique in a particular area (e.g., when there are no other landmarks, or no other landmarks of the same type), the sparse map may use data indicating the type of the landmark (the landmark or the particular type of landmark), and during navigation (e.g., autonomous navigation) when a camera mounted on an autonomous vehicle captures an image of an area containing the landmark (or the particular type of landmark), the processor may process the image, detect the landmark (if it actually exists in the image), classify the image as a landmark (or as a particular type of landmark), and correlate the position of the image with the position of the landmark stored in the sparse map.

[0196] Generation of Sparse Map

[0197] In some embodiments, the sparse map may include at least one line representation of road surface features extending along a road segment and a plurality of landmarks associated with the road segment. In a particular aspect, the sparse map may be generated via "crowdsourcing", for example, via image analysis of a plurality of images acquired when one or more vehicles cross a road segment.

[0198] FIG. 8 shows a sparse map 800 that one or more vehicles, such as vehicle 200 (which may be an autonomous vehicle), may access to provide autonomous vehicle navigation. The sparse map 800 may be stored in a memory such as memory 140 or 150. Such a memory device may include any type of non-transitory storage device or computer-readable medium. For example, in some embodiments, memory 140 or 150 may include a hard drive, a compact disk, a flash memory, a magnetic-based memory device, an optical-based memory device, etc. In some embodiments, the sparse map 800 may be stored in a database (e.g., map database 160) that may be stored in memory 140 or 150, or in another type of storage device.

[0199] In some embodiments, the sparse map 800 may be stored in a storage device or a non-transitory computer-readable medium mounted on the vehicle 200 (e.g., a storage device included in a navigation system mounted on the vehicle 200). A processor mounted on the vehicle 200 (e.g., the processing unit 110) may access the sparse map 800 stored in a storage device or a computer-readable medium mounted on the vehicle 200 to generate navigation instructions for guiding the autonomous vehicle 200 when the vehicle crosses a road segment.

[0200] 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 a computer-readable medium provided on a remote server that communicates with the vehicle 200 or a device associated with the vehicle 200. A processor mounted on the vehicle 200 (e.g., the processing unit 110) may receive data included in the sparse map 800 from the remote server and execute the data for guiding the autonomous driving of the vehicle 200. In such an embodiment, the remote server may store all or only a part of the sparse map 800. Accordingly, a storage device or a computer-readable medium mounted on the vehicle 200 and / or one or more additional vehicles may store the remaining portion of the sparse map 800.

[0201] Furthermore, in such an embodiment, the sparse map 800 may be accessible to a plurality of vehicles (e.g., dozens, hundreds, thousands, or millions of vehicles, etc.) traversing various road segments. It should also be noted that the sparse map 800 may include a plurality of submaps. For example, in some embodiments, the sparse map 800 may include hundreds, thousands, millions, 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 road may access any number of local maps related to the location where the vehicle is traveling. The local map areas of the sparse map 800 may be stored together with a global navigation satellite system (GNSS) key as an index to the database of the sparse map 800. Thus, the calculation of the steering angle for navigating the host vehicle in this system can be performed without relying on the GNSS position, road features, or landmarks of the host vehicle, although such GNSS information may be used to search for the relevant local map.

[0202] In general, a sparse map 800 can be generated based on data collected from one or more vehicles as the one or more vehicles travel along a road. For example, sensors (e.g., cameras, speedometers, GPS, accelerometers, etc.) mounted on the one or more vehicles can be used to record the trajectories of the one or more vehicles as they travel along the road, and a polynomial representation of a preferred trajectory for a vehicle traveling subsequently along the road can be determined based on the collected trajectories traveled by the one or more vehicles. Similarly, data collected by the one or more vehicles can assist in the identification of potential landmarks along a particular road. Data collected from traversing vehicles can also be used to identify road profile information such as road width profile, road roughness profile, lane spacing profile, road conditions, etc. Using the information collected, a sparse map 800 is generated and can be distributed (e.g., for local storage or via on-the-fly data transmission) for use in navigating one or more autonomous vehicles. However, in some embodiments, map generation may not end at the initial generation of the map. As discussed in more detail below, the sparse map 800 can be continuously or periodically updated based on data collected from vehicles as those vehicles continue to traverse the roads included in the sparse map 800.

[0203] The data recorded in the sparse map 800 may include location information based on Global Positioning System (GPS) data. For example, the location information may be included in the sparse map 800 of various map elements, including, for example, landmark locations, road profile locations, and the like. The locations of the map elements included in the sparse map 800 can be obtained using GPS data collected from a vehicle crossing a road. For example, a vehicle passing through an identified landmark can determine the location of the identified landmark using the GPS location information associated with the vehicle and can determine the location of the identified landmark relative to the vehicle (e.g., based on image analysis of data collected from one or more cameras mounted on the vehicle). Such determination of the location of the identified landmark (or other features included in the sparse map 800) can be repeated when additional vehicles pass through the location of the identified landmark. Some or all of the additional determinations can be used to refine the location information stored in the sparse map 800 in relation to the identified landmark. For example, in some embodiments, multiple position measurements associated with a particular feature stored in the sparse map 800 can be averaged together. However, any other mathematical operation can also be used to refine the stored location of a map element based on multiple determined locations of the map element.

[0204] The sparse map of the disclosed embodiments can enable autonomous navigation of a vehicle using a relatively small amount of stored data. In some embodiments, the sparse map 800 can have a data density (e.g., including data representing target trajectories, landmarks, and other stored road features) 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. In some embodiments, the data density of the sparse map 800 can be less than 10 KB per kilometer of road, or less than 2 KB per kilometer of road (e.g., 1.6 KB per kilometer), or less than or equal to 10 KB per kilometer of road, or less than or equal to 20 KB per kilometer of road. In some embodiments, most, if not all, of the roads in the United States can be autonomously navigated using a sparse map having a total data of 4 GB or less. These data density values can represent an average across the entire sparse map 800, across local maps within the sparse map 800, and / or across specific road segments within the sparse map 800.

[0205] As described above, the sparse map 800 may include a representation of a plurality of target trajectories 810 for guiding autonomous driving or navigation along a road segment. Such target trajectories may be stored as 3D splines. The target trajectories stored in the sparse map 800 may be determined, for example, based on two or more reconstructed trajectories of previous crossings of a vehicle along a particular road segment. A road segment may be associated with a single target trajectory or a plurality of target trajectories. For example, in a two-lane road, a first target trajectory may be stored to represent an intended driving route along the road in a first direction, and a second target trajectory may be stored to represent an intended driving route along the road in another direction (e.g., the opposite direction of the first direction). Additional target trajectories may be stored for a particular road segment. For example, in a multi-lane road, one or more target trajectories representing the intended driving routes of vehicles in one or more lanes associated with the multi-lane road may be stored. In some embodiments, each lane of a multi-lane road may be associated with its own target trajectory. In other embodiments, there may be fewer stored target trajectories than there are lanes present in the multi-lane road. In such a case, a vehicle navigating a multi-lane road may use any of the stored target trajectories to guide navigation taking into account the amount of lane offset from the lane in which the target trajectory is stored (e.g., if a vehicle is traveling in the leftmost lane of a three-lane highway and the target trajectory is stored only for the center lane of the highway, the vehicle may use the target trajectory of the center lane to navigate, taking into account the amount of lane offset between the center lane and the leftmost lane, when generating navigation instructions).

[0206] In some embodiments, the target trajectory may represent the ideal path that the vehicle should take when driving. The target trajectory may be disposed, for example, substantially at the center of the driving lane. In other cases, the target trajectory may be disposed at other locations with respect 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, etc. In such cases, the navigation based on the target trajectory may include a determined amount of offset to be maintained with respect to the position of the target trajectory. Further, in some embodiments, the determined amount of offset to be maintained with respect to the position of the target trajectory may vary based on the type of vehicle (e.g., a passenger car including two axles may have a different offset along at least a portion of the target trajectory than a truck including three or more axles).

[0207] The sparse map 800 may also include data related to a plurality of predetermined landmarks 820 associated with a particular road segment, local map, etc. As discussed in more detail below, these landmarks can be used to navigate the autonomous vehicle. For example, in some embodiments, the landmarks can be used to determine the current position of the vehicle relative to the stored target trajectory. Using this position information, the autonomous vehicle may be able to adjust its direction of travel to match the direction of the target trajectory at the determined position.

[0208] The plurality of landmarks 820 may be identified at any suitable interval and stored in the sparse map 800. In some embodiments, the landmarks may be stored at a relatively high density (e.g., every few meters or more). However, in some embodiments, significantly large landmark interval values may be used. For example, in the sparse map 800, the identified (or recognized) landmarks may be separated by intervals of 10 meters, 20 meters, 50 meters, 100 meters, 1 kilometer, or 2 kilometers. In some cases, the identified landmarks may be located more than 2 kilometers apart.

[0209] While determining the landmark pairs and thus the vehicle position relative to the target trajectory, the vehicle may navigate based on dead reckoning in which the vehicle uses sensors to determine its own movement and estimates its position relative to the target trajectory. Since errors can accumulate during dead reckoning navigation, the accuracy of the positioning relative to the target trajectory may gradually decrease over time. The vehicle may use landmarks present in the sparse map 800 (and their known positions) to remove errors induced by dead reckoning in positioning. In this way, the identified landmarks included in the sparse map 800 may function as navigation anchors from which the exact position of the vehicle relative to the target trajectory can be determined. In positioning, since a certain degree of error may be tolerated, it is not necessary for the identified landmarks to always be available to the autonomous vehicle. Rather, as described above, appropriate navigation may be possible based on landmark intervals of 10 meters, 20 meters, 50 meters, 100 meters, 500 meters, 1 kilometer, 2 kilometers, or more. In some embodiments, the density of one identified landmark per 1 km of road may be sufficient to maintain a longitudinal positioning accuracy within 1 m. Thus, it is not always necessary to store all potential landmarks that appear along the road segment in the sparse map 800.

[0210] Furthermore, in some embodiments, lane marks may be used for vehicle localization between landmark intervals. By using lane marks between landmark intervals, accumulation during dead reckoning navigation can be minimized.

[0211] In addition to the target orbit and the identified landmarks, the sparse map 800 may include information related to various other road features. For example, FIG. 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 3D polynomial descriptions on the left and right sides of the road. Such polynomials representing the left and right sides of a single lane are shown in FIG. 9A. Regardless of the number of lanes a road may have, the road can be represented using polynomials in a similar manner as shown in FIG. 9A. For example, the left and right sides of a multi-lane road can be represented by polynomials similar to those shown in FIG. 9A, and the markings of the intermediate lanes included in the multi-lane road (e.g., dashed markings representing lane boundaries, solid yellow lines representing boundaries between lanes traveling in different directions, etc.) can also be represented using polynomials as shown in FIG. 9A.

[0212] As shown in FIG. 9A, lane 900 can be represented using polynomials (e.g., first-degree, second-degree, third-degree, or polynomials of any suitable degree). For the sake of illustration, lane 900 is shown as a 2D lane and the polynomials are shown as 2D polynomials. As shown in FIG. 9A, lane 900 includes a left side 910 and a right side 920. In some embodiments, multiple polynomials can be used to represent the positions on each side of the boundary of a road or lane. For example, each of the left side 910 and the right side 920 can be represented by a plurality of polynomials of any suitable length. In some cases, the polynomials can be about 100 m in length, but other lengths longer or shorter than 100 m can also be used. Further, the polynomials can be overlapped with each other to facilitate seamless transitions when navigating based on the polynomials that will be encountered later as the host vehicle travels along the road. For example, each of the left side 910 and the right side 920 can be represented by a plurality of third-degree polynomials that are separated into segments of about 100 meters (an example of a first predetermined range) and overlap each other by about 50 meters. The polynomials representing the left side 910 and the right side 920 can be in the same order or not in the same order. For example, in some embodiments, some of the polynomials can be second-degree polynomials, some can be third-degree polynomials, and some can be fourth-degree polynomials.

[0213] In the example shown in FIG. 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 but follow the positions on each side of the road. Polynomial segments 911, 912, 913, 914, 915, and 916 are approximately 100 meters in length and overlap with adjacent segments within a series by approximately 50 meters. However, as described above, the length and the amount of overlap can also use different polynomials. For example, the polynomials can be 500m, 1km, or longer in length, and the amount of overlap can vary from 0 - 50m, 50m - 100m, or more than 100m. Further, FIG. 9A is shown as representing polynomials that extend in a 2D space (e.g., on the surface of a paper), and it should be understood that these polynomials can represent curves that extend in three dimensions (e.g., including a height component) to represent changes in elevation of the road segment in addition to the X - Y curvature. In the example shown in FIG. 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.

[0214] Returning to the target trajectory of the sparse map 800, FIG. 9B shows a cubic polynomial representing the target trajectory of a vehicle traveling along a particular road segment. The target trajectory represents not only the X - Y path that the host vehicle should travel along the particular road segment, but also the change in elevation that the host vehicle experiences as it travels along the road segment. Thus, each target trajectory within the sparse map 800 can be represented by one or more cubic polynomials, such as the cubic polynomial 950 shown in FIG. 9B. The sparse map 800 can include a plurality (e.g., millions or billions or more to represent the trajectories of vehicles along various road segments along roads around the world) of trajectories. In some embodiments, each target trajectory can correspond to a spline connecting cubic polynomial segments.

[0215] Regarding the data footprint of polynomial curves stored in the sparse map 800, in some embodiments, each cubic polynomial is represented by four parameters, each of which may require 4 bytes of data. A suitable representation can be obtained with cubic polynomials that require approximately 192 bytes of data per 100 m. This can be rephrased as approximately 200 KB per hour in terms of the data usage / transfer requirements of a host vehicle traveling at approximately 100 km / hr.

[0216] The sparse map 800 may describe the lane network using a combination of geometric descriptors and metadata. The geometry may be described by polynomials or splines as described above. The metadata may describe the number of lanes, special characteristics (such as carpool lanes), and optionally other sparse labels. The total footprint of such indicators can be very small.

[0217] Thus, a sparse map according to an embodiment of the present disclosure may include at least one line representation of a road surface feature extending along a road segment, and each line representation represents a path along the road segment that substantially corresponds to the road surface feature. In some embodiments, as discussed above, at least one line representation of the road surface feature may include a spline, a polynomial representation, or a curve. Further, in some embodiments, the road surface feature may include at least one of a road edge or a lane mark. Further, as discussed below with respect to "crowdsourcing", the road surface feature may be identified by image analysis of a plurality of images obtained when one or more vehicles cross the road segment.

[0218] As described above, the sparse map 800 may include a plurality of predetermined landmarks associated with the road segment. Instead of storing actual images of the landmarks and relying, for example, on image recognition analysis based on the captured and stored images, each landmark in the sparse map 800 can be represented and recognized using less data than is required for the actual images to be stored. The data representing the landmarks may include sufficient information to describe or identify the landmarks along the road. By storing data describing the characteristics of the landmarks rather than the actual images of the landmarks, the size of the sparse map 800 can be reduced.

[0219] FIG. 10 shows an example of a type of landmark that can be represented by a sparse map 800. Landmarks can include visible and distinguishable objects along a road segment. The landmarks can be fixed and selected so as not to be frequently changed with respect to location and / or content. The landmarks included in the sparse map 800 can be useful for determining the position of the vehicle 200 relative to the target trajectory when the vehicle crosses a particular road segment. Examples of landmarks can include traffic signs, direction signs, general signs (e.g., rectangular signs), roadside equipment (e.g., street light poles, reflectors, etc.), and other suitable categories. In some embodiments, lane markings on the road can also be included as landmarks in the sparse map 800.

[0220] The example of landmarks shown in FIG. 10 includes traffic signs, direction signs, roadside equipment, and general signs. Traffic signs can 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 signs (e.g., traffic signal sign 1015), stop signs (e.g., stop sign 1020). Direction signs can include signs that include one or more arrows indicating one or more directions to different locations. For example, direction signs can include highway signs 1025 having arrows for directing the vehicle in the direction of different roads or locations, exit signs 1030 having arrows for directing the vehicle in the direction of exiting the road, etc. Thus, at least one of the plurality of landmarks can include a road sign.

[0221] General signs can be unrelated to traffic. For example, general signs can include billboards used for advertising, or welcome boards adjacent to the boundaries between two countries, states, counties, cities, or towns. FIG. 10 shows a general sign 1040 (“Joe's Restaurant”). As shown in FIG. 10, the general sign 1040 can have a rectangular shape, but the general sign 1040 can have other shapes such as square, circular, triangular, etc.

[0222] Landmarks can also include roadside equipment. The roadside equipment can be an object that is not a sign and may not be related to traffic or direction. For example, the roadside equipment can include street light poles (e.g., street light pole 1035), power line poles, signal poles, etc.

[0223] Landmarks can also include beacons that are specially designed for use in the navigation system of autonomous vehicles. For example, such beacons can include stand-alone structures that are arranged at predetermined intervals to assist the navigation of the host vehicle. Such beacons can also include visual / graphical information that is added to existing road signs (e.g., icons, emblems, barcodes, etc.) that can be identified or recognized by vehicles traveling along a road segment. Such beacons can also include electronic components. In such embodiments, electronic beacons (e.g., RFID tags, etc.) can be used to transmit non-visual information to the host vehicle. Such information can include, for example, landmark identification and / or landmark position information that can be used when the host vehicle determines its position along a target trajectory.

[0224] In some embodiments, the landmarks included in the sparse map 800 may be represented by data objects of a predetermined size. The data representing the landmark may include any suitable parameters for identifying a particular landmark. For example, in some embodiments, the landmarks stored in the sparse map 800 include the physical size of the landmark (e.g., to support the estimation of the distance to the landmark based on a known size / scale), the distance to the previous landmark, the lateral offset, the height, the type code (e.g., the type of landmark - direction sign, traffic sign, etc.), the GPS coordinates (e.g., to support global localization), and other suitable parameters. Each parameter may be associated with a data size. For example, the landmark size may be stored using 8 bytes of data. The distance to the previous landmark, the lateral offset, and the height may be specified using 12 bytes of data. The type code associated with a landmark such as a direction sign or a traffic sign may require approximately 2 bytes of data. In the case of a general sign, the image signature that enables the identification of the general sign may be stored using 50 bytes of data storage. The GPS position of the landmark may be associated with 16 bytes of data storage. These data sizes for each parameter are merely examples, and other data sizes may also be used.

[0225] Representing landmarks in a sparse map 800 in this way can provide an efficient and waste-free solution for representing landmarks in a database. In some embodiments, signs can be referred to as semantic signs and non-semantic signs. Semantic signs can include any class of signs with a standardized meaning (e.g., speed limit signs, warning signs, direction signs, etc.). Non-semantic signs can include any signs not associated with a standardized meaning (e.g., general advertising signs, signs identifying a business, etc.). For example, each semantic sign can be represented by 38 bytes of data (e.g., 8 bytes for size, distance to the previous landmark, lateral offset, 12 bytes for height, 2 bytes for type code, 16 bytes for GPS coordinates). The sparse map 800 can use a tag system to represent landmark types. In some cases, each traffic sign or direction sign can be associated with a unique tag and stored in the database as part of the landmark ID. For example, the database can include about 1000 different tags to represent various traffic signs and about 10000 different tags to represent direction signs. Of course, any appropriate number of tags can be used and additional tags can be created as needed. General-purpose signs can be represented using less than about 100 bytes in some embodiments (e.g., about 86 bytes including 8 bytes for size, distance to the previous landmark, lateral offset, and 12 bytes for height, 50 bytes for image signature, and 16 bytes for GPS coordinates).

[0226] Thus, in the case of semantic road signs that do not require an image signature, the impact on the data density of the sparse map 800 can be approximately 760 bytes per kilometer, even with a relatively high landmark density of about 1 per 50 m (e.g., 20 landmarks per kilometer x 38 bytes per landmark = 760 bytes). Even in the case of general-purpose signs that include an image signature component, the impact on the data density is approximately 1.72 KB per kilometer (e.g., 20 landmarks per kilometer x 86 bytes per landmark = 1,720 bytes). In the case of semantic road signs, this corresponds to approximately 76 KB of data usage per hour for a vehicle traveling at 100 km / hr. In the case of general-purpose signs, this corresponds to approximately 170 KB per hour for a vehicle traveling at 100 km / hr.

[0227] In some embodiments, generally rectangular objects, such as rectangular signs, can be represented within the sparse map 800 with data of 100 bytes or less. The representation of generally rectangular objects (e.g., general sign 1040) in the sparse map 800 can include a condensed image signature (e.g., condensed image signature 1045) associated with the generally rectangular object. This condensed image signature can be used, for example, as a recognized landmark to assist in the identification of general-purpose signs. Such a condensed image signature (e.g., image information derived from the actual image data representing the object) can avoid the need to store the actual image of the object or perform comparative image analysis on the actual image to recognize the landmark.

[0228] Referring to FIG. 10, the sparse map 800 may include or store a condensed image signature 1045 associated with the general label 1040, rather than an actual image of the general label 1040. For example, after an image capture device (e.g., image capture devices 122, 124, or 126) captures an image of the general label 1040, a processor (e.g., image processor 190, or any other processor capable of processing the image either on the host vehicle or remotely located) may perform image analysis to extract / create a condensed image signature 1045 that includes a unique signature or pattern associated with the general label 1040. In one embodiment, the condensed image signature 1045 may include a shape, color pattern, brightness pattern, or any other feature that can be extracted from an image of the general label 1040 to describe the general label 1040.

[0229] For example, in FIG. 10, the circles, triangles, and stars shown in the condensed image signature 1045 may represent regions of different colors. The patterns represented by the circles, triangles, and stars may be stored in the sparse map 800, for example, within 50 bytes specified to include an 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 are intended to conceptually represent recognizable regions having other irregularities of distinguishable color differences, text regions, graphic shapes, or characteristics that can be associated with general labels. Such condensed image signatures can be used to identify landmarks in the form of general labels. For example, the condensed image signature can be used to perform a same-or-not analysis based on a comparison between, for example, image data captured using a camera mounted on an autonomous vehicle and the stored condensed image signature.

[0230] Accordingly, the plurality of landmarks can be identified by image analysis of a plurality of images obtained when one or more vehicles cross a road segment. As will be described below with respect to "crowdsourcing", in some embodiments, the image analysis for identifying the plurality of landmarks can include accepting potential landmarks when the ratio of images in which a landmark appears to images in which the landmark does not appear exceeds a threshold. Further, in some embodiments, the image analysis for identifying the plurality of landmarks can include rejecting potential landmarks when the ratio of images in which the landmark does not appear to images in which the landmark appears exceeds a threshold.

[0231] Returning to the target trajectory that the host vehicle can use to navigate a particular road segment, FIG. 11A shows a polynomial representation trajectory captured during the process of constructing 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 of previous crossings 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 aggregation of two or more reconstructed trajectories of previous crossings 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 of previous crossings of the vehicle along the same road segment. Other mathematical operations can also be used to construct the target trajectory along the road path based on the reconstructed trajectories collected from vehicles crossing along the road segment.

[0232] As shown in FIG. 11A, the road segment 1100 can be traveled by a plurality of vehicles 200 at different times. Each vehicle 200 can collect data related to the route taken by the vehicle along the road segment. The route traveled by a particular vehicle can be determined based on, among other potential information sources, camera data, accelerometer information, speed sensor information, and / or GPS information. Such data can be used to reconstruct the trajectories of vehicles traveling along the road segment, and based on these reconstructed trajectories, the target trajectory (or trajectories) of a particular road segment can be determined. Such a target trajectory can represent the preferred route of the host vehicle (e.g., guided by an autonomous navigation system) when the vehicle travels along the road segment.

[0233] In the example shown in FIG. 11A, the first reconstructed trajectory 1101 can be determined based on data received from a first vehicle crossing the road segment 1100 during a first period (e.g., the first day), the second reconstructed trajectory 1102 can be obtained from a second vehicle crossing the road segment 1100 during a second period (e.g., the second day), and the third reconstructed trajectory 1103 can be obtained from a third vehicle crossing the road segment 1100 during a third period (e.g., the third day). Each of the trajectories 1101, 1102, and 1103 can be represented by a polynomial such as a three-dimensional polynomial. Note that in some embodiments, any of the reconstructed trajectories can be provided to and assembled by a vehicle crossing the road segment 1100.

[0234] Additionally or alternatively, such a reconstructed trajectory can be determined on the server side based on information received from a vehicle traversing road segment 1100. For example, in some embodiments, vehicle 200 can transmit data related to their movement along road segment 1100 (e.g., among other things, steering angle, direction of travel, time, position, speed, detected road geometry, and / or detected landmarks, etc.) to one or more servers. The server can reconstruct the trajectory of vehicle 200 based on the received data. The server can also generate a target trajectory for guiding the navigation of an autonomous vehicle that will later travel along the same road segment 1100 based on the first, second, and third trajectories 1101, 1102, and 1103. The target trajectory can be associated with a single previous traversal of the road segment, but in some embodiments, each target trajectory included in the sparse map 800 can be determined based on two or more reconstructed trajectories of vehicles traversing the same road segment. In FIG. 11A, the target trajectory is represented by 1110. In some embodiments, the target trajectory 1110 can be generated based on the average of the first, second, and third trajectories 1101, 1102, and 1103. In some embodiments, the target trajectory 1110 included in the sparse map 800 can be an aggregation (e.g., a weighted combination) of two or more reconstructed trajectories.

[0235] FIGS. 11B and 11C further illustrate the concept of a target trajectory associated with a road segment existing within geographical area 1111. As shown in FIG. 11B, a first road segment 1120 within geographical area 1111 can include a multi-lane road including 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 the first direction. Lanes 1122 and 1124 can be separated by a double yellow line 1123. Geographical area 1111 can also include a branch road segment 1130 that intersects road segment 1120. Road segment 1130 can include a two-lane road, with each lane designated for a different direction of travel. Geographical area 1111 can also include other road features such as a stop line 1132, a stop sign 1134, a speed limit sign 1136, and a hazard sign 1138, etc.

[0236] As shown in FIG. 11C, the sparse map 800 may include a local map 1140 that includes a road model for assisting the autonomous navigation of vehicles within the geographical area 1111. For example, the local map 1140 may include the target trajectories of one or more lanes associated with the road segments 1120 and / or 1130 within the geographical area 1111. For example, the local map 1140 may include the target trajectories 1141 and / or 1142 that the autonomous vehicle may access or depend on when crossing the lane 1122. Similarly, the local map 1140 may include the target trajectories 1143 and / or 1144 that the autonomous vehicle may access or depend on when crossing the lane 1124. Further, the local map 1140 may include the target trajectories 1145 and / or 1146 that the autonomous vehicle may access or depend on when crossing the road segment 1130. The target trajectory 1147 represents the preferred path that the autonomous vehicle should follow when transitioning from the lane 1120 (specifically, corresponding to the target trajectory 1141 associated with the rightmost lane of the lane 1120) to the road segment 1130 (specifically, corresponding to the target trajectory 1145 associated with the first side of the road segment 1130). Similarly, the target trajectory 1148 represents the preferred path that the autonomous vehicle should follow when transitioning from the road segment 1130 (specifically, corresponding to the target trajectory 1146) to a part of the road segment 1124 (specifically, as shown, corresponding to the target trajectory 1143 associated with the left lane of the lane 1124).

[0237] The sparse map 800 may also include representations of other road-related features associated with the geographic region 1111. For example, the sparse map 800 may also include representations of one or more landmarks identified in the geographic region 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 such that the vehicle can adjust its direction of travel to match the direction of the target trajectory at the determined position.

[0238] In some embodiments, the sparse map 800 may also include a road signature profile. Such a road signature profile may be associated with an identifiable / measurable variation of at least one parameter associated with the road. For example, in some cases, such a profile may be associated with variations in road surface roughness for a particular road segment, variations in road width across a particular road segment, variations in the distance between dashed lines drawn along a particular road segment, variations in the curvature of the road along a particular road segment, etc. FIG. 11D shows an example of a road signature profile 1160. The profile 1160 may represent any of the above parameters, or other parameters, but in one example, the profile 1160 may represent a measure of road surface roughness obtained, for example, by monitoring one or more sensors that provide an output indicative of the amount of suspension displacement when a vehicle travels a particular road segment.

[0239] Alternatively, or simultaneously, profile 1160 may represent a change in road width determined based on image data obtained via a camera mounted on a vehicle traveling on a particular road segment. Such a profile may be useful, for example, in determining a particular position of an autonomous vehicle relative to a particular target trajectory. That is, when an autonomous vehicle crosses a road segment, the autonomous vehicle may measure a profile associated with one or more parameters associated with the road segment. If the measured profile can be correlated / matched with a predetermined profile that plots the change in the parameter with respect to the position along the road segment, the current position along the road segment, and thus the current position relative to the target trajectory of the road segment, can be determined (e.g., by overlapping corresponding regions of the predetermined profile being measured).

[0240] In some embodiments, 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. A sparse map 800 including such different trajectories may be provided to different autonomous vehicles of different users. For example, some users may prefer to avoid toll roads, while other users may prefer to take the shortest or fastest route regardless of whether there are toll roads on the route. The disclosed system may generate different sparse maps with different trajectories based on such different user preferences or profiles. As another example, some users may prefer to drive in lanes that move at high speeds, while other users may prefer to always maintain a position in the center lane.

[0241] Different trajectories can be generated based on different environmental conditions such as day and night, snow, rain, fog, etc., and can be included in the sparse map 800. An autonomous vehicle traveling under different environmental conditions can provide the sparse map 800 generated based on such different environmental conditions. In some embodiments, a camera provided to the autonomous vehicle can detect environmental conditions and provide such information to a server that generates and provides a sparse map. For example, the server can generate or update the already generated sparse map 800 to include trajectories that may be more suitable or safer for autonomous driving under the detected environmental conditions. The update of the sparse map 800 based on environmental conditions can be performed dynamically when the autonomous vehicle is traveling along the road.

[0242] Other different parameters related to driving can also be used as a basis for generating and providing different sparse maps to different autonomous vehicles. For example, when an autonomous vehicle is traveling at a high speed, turning can become difficult. When an autonomous vehicle is proceeding along a specific trajectory, a trajectory associated with a specific lane rather than the road can be included in the sparse map 800 so that the vehicle can maintain within the specific lane. When an image captured by a camera mounted on the autonomous vehicle indicates that the vehicle has drifted outside the lane (e.g., crossed the lane mark), an operation can be triggered inside the vehicle to return the vehicle to the designated lane according to a specific trajectory.

[0243] Cloud sourcing of sparse maps

[0244] In some embodiments, the disclosed systems and methods may generate a sparse map for autonomous vehicle navigation. For example, the disclosed systems and methods may use crowdsourced data to generate a sparse map that can be used for one or more autonomous vehicles to navigate along a road system. As used herein, "crowdsourcing" means receiving data from various vehicles (e.g., autonomous vehicles) traveling on road segments at different times and using such data to generate and / or update a road model. The model may then be transmitted to vehicles traveling along the road segment or other vehicles later to assist in the navigation of autonomous vehicles. The road model may include a plurality of target trajectories representing preferred trajectories for the vehicle to follow when crossing a road segment. The target trajectories may be the same as the reconstructed actual trajectories collected from vehicles crossing the road segment and may be transmitted from the vehicle to the server. In some embodiments, the target trajectories may be different from the actual trajectories previously taken when one or more vehicles cross the road segment. The target trajectories may be generated based on the actual trajectories (e.g., by averaging or other suitable operations).

[0245] The vehicle trajectory data that a vehicle may upload to the server may correspond to, be based on, or be related to the actual reconstructed trajectory of the vehicle, but may be different from the actual reconstructed trajectory, and may correspond to a recommended trajectory. For example, a vehicle may modify the actual reconstructed trajectory and transmit (e.g., recommend) the modified actual trajectory to the server. The road model may use the recommended modified trajectory as the target vehicle trajectory for the autonomous navigation of other vehicles.

[0246] In addition to trajectory information, other information potentially used in 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 positions of the landmarks. The landmarks may be used by the autonomous vehicle's navigation system to determine and / or adjust the position of the vehicle along the target trajectory.

[0247] When the vehicle is traveling along a road, the reconstructed trajectory that the vehicle can generate can be obtained by any suitable method. In some embodiments, the reconstructed trajectory can be developed by, for example, using self-motion estimation (e.g., camera, and thus the three-dimensional translation and three-dimensional rotation of the vehicle body) to piece together segments of the vehicle's motion. The estimation of rotation and translation can be determined based on the analysis of images captured by one or more image capture devices, along with information from other sensors or devices such as inertial sensors and speed sensors. For example, the inertial sensor can include an accelerometer or other suitable sensors configured to measure changes in the translation and / or rotation of the vehicle body. The vehicle can include a speed sensor that measures the speed of the vehicle.

[0248] In some embodiments, the self-motion of the camera (and thus the vehicle body) can be estimated based on the optical flow analysis of the captured images. The optical flow analysis of a series of images identifies the movement of pixels from the series of images and determines the movement of the vehicle based on the identified movement. The self-motion is integrated over time along the road segment, and the trajectory associated with the road segment traversed by the vehicle can be reconstructed.

[0249] Data (e.g., reconstructed trajectories) collected by multiple vehicles during multiple runs along a road segment at different times can be used to construct a road model (e.g., including a target trajectory, etc.) included in the sparse data map 800. To improve the accuracy of the model, the data collected by multiple vehicles during multiple runs along a road segment at different times can also be averaged. In some embodiments, data regarding the geometry and / or landmarks of the road can be received from multiple vehicles traveling on a common road segment at different times. Such data received from different vehicles can be combined to generate and / or update the road model.

[0250] The geometry of the reconstructed track (and target track) along the road segment can be represented by a curve in 3D space, which can be a spline connecting 3D polynomials. The reconstructed track curve can be determined from the analysis of a plurality of images captured by a video stream or a camera attached to the vehicle. In some embodiments, the position is identified in each frame or image several meters ahead of the current position of the vehicle. This location is where the vehicle is expected to travel within a given period of time. This operation can be repeated for each frame, and at the same time, the vehicle can calculate its own motion (rotation and translation) of the camera. In each frame or image, a short-distance model of the desired path is generated by the vehicle within the reference frame attached to the camera. By stitching together the short-distance models, a 3D model of the road within a coordinate frame, which can be any coordinate frame or a predetermined coordinate frame, can be obtained. The 3D model of the road can then be fitted by a spline that can include or connect one or more polynomials of an appropriate degree.

[0251] One or more detection modules can be used to conclude the short-distance road model in each frame. For example, a bottom-up lane detection module can be used. The bottom-up lane detection module can be useful when lane markings are drawn on the road. This module can search for the edges in the image and assemble them together to form lane markings. A second module can be used together with the bottom-up lane detection module. The second module is an end-to-end deep neural network that can be trained to predict the correct short-distance path from the input image. In either module, the road model is detected in the image coordinate frame and can be converted to a 3D space that can be virtually connected to the camera.

[0252] The reconstructed trajectory modeling method may introduce error accumulation through the integration of long-term self-motion that may include noise components, but such errors may not be significant because the generated model can provide sufficient accuracy for local-scale navigation. Additionally, external information sources such as satellite images or geodetic measurements can be used to cancel the integration error. For example, the disclosed systems and methods may use a GNSS receiver to cancel the accumulated error. However, GNSS positioning signals are not always available and accurate. The disclosed systems and methods may enable steering applications that are weakly dependent on the availability and accuracy of GNSS positioning. In such systems, the use of GNSS signals may be restricted. For example, in some embodiments, the disclosed system may use GNSS signals only for the purpose of database indexing.

[0253] In some embodiments, the range scale (e.g., local scale) that may be relevant to an autonomous vehicle navigation steering application may be on the order of 50 meters, 100 meters, 200 meters, 300 meters, etc. Geometric road models may use such distances because they are used primarily for two purposes: planning the forward trajectory and determining the vehicle's position on the road model. In some embodiments, when the control algorithm steers the vehicle according to a target point located 1.3 seconds ahead (or any other time such as 1.5 seconds, 1.7 seconds, 2 seconds, etc.), the planning task may use the model over a typical range of 40 meters ahead (or other suitable forward distances such as 20 meters, 30 meters, 50 meters, etc.). In the positioning task, the road model is used over a typical range of 60 meters behind the vehicle (or other suitable distances such as 50 meters, 100 meters, 150 meters, etc.) according to a method called "tail alignment" that will be described in more detail in another section. The disclosed systems and methods may generate a geometric model with sufficient accuracy over a specific range such as 100 meters so that the planned trajectory does not deviate by more than 30 cm from the lane center, for example.

[0254] As described above, the 3D road model can be constructed by detecting short-distance sections and connecting them together. Connecting them together can be enabled by calculating a six-degree-of-freedom ego-motion model using video and / or images captured by a camera, data from an inertial sensor reflecting the movement of the vehicle, and the speed signal of the host vehicle. The cumulative error can be small enough at some local range scales, such as about 100 meters. Throughout this range scale, a particular road segment can be completed in a single run.

[0255] In some embodiments, multiple runs can be used to average the resulting model and further improve its accuracy. The same vehicle can drive the same route multiple times, or the model data collected by multiple vehicles can be sent to a central server. In either case, a matching procedure can be performed to identify overlapping models and enable averaging to generate a target trajectory. The constructed model (e.g., including the target trajectory) can be used for steering when a convergence criterion is met. Subsequent runs can be used to further improve the model and to accommodate infrastructure changes.

[0256] When multiple vehicles are connected to a central server, it becomes possible to share driving experiences (such as detection data) among the multiple vehicles. Each vehicle client can store a partial copy of the universal road model that may be relevant to its current position. A two-way update procedure between the vehicle and the server can be executed by the vehicle and the server. The concept of the small footprint discussed above enables the disclosed systems and methods to perform two-way updates using a very narrow bandwidth.

[0257] Information related to potential landmarks can also be determined and transferred to the central server. For example, the disclosed systems and methods can determine one or more physical characteristics of potential landmarks based on one or more images including the landmarks. Physical characteristics can include the physical size of the landmark (e.g., height, width), the distance from the vehicle to the landmark, the distance from the landmark to the previous landmark, 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 the landmark, the identification of text on the landmark, etc. For example, the vehicle can analyze one or more images captured by a camera to detect potential landmarks such as speed limit signs.

[0258] The vehicle can determine the distance from the vehicle to the landmark based on the analysis of one or more images. In some embodiments, the distance can be determined based on the analysis of the image of the landmark using appropriate image analysis methods such as scaling methods and / or optical flow methods. In some embodiments, the disclosed systems and methods can be configured to determine the type or classification of potential landmarks. If the vehicle determines that a particular potential landmark corresponds to a predetermined type or classification stored in the sparse map, it may be sufficient for the vehicle to communicate the display of the type or classification of the landmark along with its position to the server. The server can store such a display. Later, other vehicles can capture an image of the landmark, process the image (e.g., using a classifier), and compare the result of processing the image with the display regarding the type of the landmark stored in the server. There can be various types of landmarks, and different types of landmarks can be associated with different types of data uploaded and stored in the server, and through different processes installed in the vehicle, the landmark can be detected and information about the landmark can be transmitted to the server, and the system installed in the vehicle can receive the landmark data from the server and use the landmark data to identify the landmark in autonomous navigation.

[0259] In some embodiments, a plurality of autonomous vehicles traveling on a road segment may communicate with a server. A vehicle (or client) may generate a curve that describes its travel in any coordinate frame (e.g., by integrating self-motion). The vehicle may detect landmarks and place them within the same frame. The vehicle may upload the curve and the landmarks to the server. The server may collect data from the vehicles over multiple trips and generate a unified road model. For example, as discussed below with respect to FIG. 19, the server may use the uploaded curves and landmarks to generate a sparse map with a unified road model.

[0260] The server may also distribute the model to clients (e.g., vehicles). For example, the server may distribute the sparse map to one or more vehicles. The server may continuously or periodically update the model when it receives new data from the vehicles. For example, the server may process the new data to evaluate whether the data contains information that should trigger an update on the server or the creation of new data. The server may distribute the updated model or the update to the vehicle to provide navigation for the autonomous vehicle.

[0261] The server may use one or more criteria to determine whether new data received from the vehicle should trigger an update of the model or the creation of new data. For example, if the new data indicates that a previously recognized landmark does not exist at a particular location or has been replaced by another landmark, the server may determine that the new data should trigger an update of the model. As another example, if the new data indicates that a road segment is closed and this is corroborated by data received from other vehicles, the server may determine that the new data should trigger an update of the model.

[0262] The server may distribute an updated model (or the updated part of the model) to one or more vehicles traveling on a road segment associated with the update to the model. The server may also distribute the updated model to a vehicle attempting to travel on a road segment associated with the update to the model, or to a vehicle of a planned trip including the 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 distribute the update or the updated model to the autonomous vehicle before the vehicle reaches the road segment.

[0263] In some embodiments, a remote server may collect trajectories and landmarks from multiple clients (e.g., vehicles traveling along a common road segment). The server may use the landmarks to match curves and create an average road model based on the trajectories collected from multiple vehicles. The server may also calculate the most likely routes at each node or junction of the road graph and road segments. For example, the remote server may align the trajectories to generate a sparse map crowdsourced from the collected trajectories.

[0264] The server can determine an arc length parameter and average landmark properties received from multiple vehicles traveling along a common road segment, such as the distance between one landmark and another (e.g., the previous landmark along a road segment) measured by the multiple vehicles, to support position identification and speed calibration along the routes of each client vehicle. The server can average the physical dimensions of landmarks measured by multiple vehicles traveling along a common road segment and recognizing the same landmark. The averaged physical dimensions can be used to support distance estimation such as the distance from a vehicle to a landmark. The server can average the lateral position of a landmark (e.g., the position from the lane in which the vehicle is traveling to the landmark) measured by multiple vehicles traveling along a common road segment and recognizing the same landmark. The averaged lateral position can be used to support lane assignment. The server can 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 can be used to support overall identification or positioning of the landmark within a road model.

[0265] In some embodiments, the server can identify model changes such as construction, detours, new signs, sign removals, etc. based on data received from vehicles. The server can update the model continuously or periodically or instantaneously when new data is received from a vehicle. The server can distribute the updated model or the updated models to vehicles to provide autonomous navigation. For example, as further discussed below, the server can use crowd-sourced data to exclude "ghost" landmarks detected by vehicles.

[0266] In some embodiments, the server may analyze a driver's intervention during autonomous driving. The server may analyze data received from the vehicle at the time and location where the intervention occurs, and / or data received before the time when the intervention occurs. The server may identify a specific portion of the data that caused or is closely related to the intervention, for example, data indicating a temporary lane closure setting, data indicating a pedestrian on the road. The server may update the model based on the identified data. For example, the server may modify one or more trajectories stored in the model.

[0267] FIG. 12 is a schematic diagram of a system that generates a sparse map using crowdsourcing (and distributes and navigates using the sparse map that is crowdsourced). FIG. 12 shows a road segment 1200 including one or more lanes. A plurality of vehicles 1205, 1210, 1215, 1220, and 1225 may travel on the road segment 1200 simultaneously or at different times (however, FIG. 12 shows them as appearing on the road segment 1200 simultaneously). At least one of the vehicles 1205, 1210, 1215, 1220, and 1225 may be an autonomous vehicle. For simplicity of this example, it is assumed that all of the vehicles 1205, 1210, 1215, 1220, and 1225 are autonomous vehicles.

[0268] Each vehicle may be similar to the vehicles disclosed in other embodiments (e.g., vehicle 200), and may include components or devices included in or associated with the vehicles disclosed in other embodiments. Each vehicle may be equipped with an image capture device or camera (e.g., image capture device 122 or camera 122). Each vehicle may communicate with a remote server 1230 through a wireless communication path 1235 via one or more networks (e.g., via a cellular network and / or the Internet, etc.) as indicated by the dotted line. Each vehicle may send data to the server 1230 and receive data from the server 1230. For example, the server 1230 may collect data from a plurality of vehicles traveling on a road segment 1200 at different times, process the collected data, and generate an autonomous vehicle road navigation model or an update to the model. The server 1230 may send the autonomous vehicle road navigation model or the update to the model to the vehicle that sent data to the server 1230. The server 1230 may send the autonomous vehicle road navigation model or the update to the model to other vehicles that will later travel on the road segment 1200.

[0269] When vehicles 1205, 1210, 1215, 1220, and 1225 travel on road segment 1200, the navigation information collected (e.g., detected, sensed, or measured) by vehicles 1205, 1210, 1215, 1220, and 1225 can be transmitted to server 1230. In some embodiments, the navigation information can be associated with a common road segment 1200. The navigation information can include trajectories associated with each of vehicles 1205, 1210, 1215, 1220, and 1225 when each vehicle travels on road segment 1200. In some embodiments, the trajectories can be reconstructed based on data detected by various sensors and devices provided to vehicle 1205. For example, the trajectory can be reconstructed based on at least one of accelerometer data, speed 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 vehicle 1205 detected by a speed sensor. Further, in some embodiments, the trajectory can be determined (e.g., by a processor mounted in each of vehicles 1205, 1210, 1215, 1220, and 1225) based on the detected ego-motion of a camera that can indicate three-dimensional translation and / or three-dimensional rotation (or rotational motion). The ego-motion of the camera (and thus the vehicle body) can be determined from the analysis of one or more images captured by the camera.

[0270] In some embodiments, the trajectory of vehicle 1205 is determined by a processor mounted in vehicle 1205 and can be transmitted to server 1230. In other embodiments, server 1230 can receive data detected by various sensors and devices mounted in vehicle 1205 and determine the trajectory based on the data received from vehicle 1205.

[0271] In some embodiments, the navigation information transmitted from vehicles 1205, 1210, 1215, 1220, and 1225 to server 1230 may include data regarding the road surface, road geometry, or road profile. The geometry of road segment 1200 may include lane structure and / or landmarks. The lane structure may include the total number of lanes in road segment 1200, the type of lanes (e.g., one-way lanes, two-way lanes, driving lanes, passing lanes, etc.), the markings on the lanes, the lane width, and the like. In some embodiments, the navigation information may include lane assignment, for example, which lane of a plurality of lanes the vehicle is driving in. For example, a numerical value "3" indicating that the vehicle is driving in the third lane from the left or right may be associated with the lane assignment. As another example, a text value "center lane" indicating that the vehicle is driving in the center lane may be associated with the lane assignment.

[0272] Server 1230 may store navigation information in a non-transitory computer-readable medium such as a hard drive, compact disk, tape, memory, etc. Server 1230 may generate at least a part of an autonomous vehicle road navigation model for a common road segment 1200 (e.g., via a processor included in Server 1230) based on navigation information received from multiple vehicles 1205, 1210, 1215, 1220, and 1225, and store the model as part of a sparse map. Server 1230 may determine trajectories associated with each lane based on cloud source 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 generate an autonomous vehicle road navigation model or a part of the model (e.g., the updated part) based on multiple trajectories determined based on cloud-sourced navigation data. Server 1230 may transmit the model or the updated part of the model to one or more of the autonomous vehicles 1205, 1210, 1215, 1220, and 1225 traveling on road segment 1200 or any other autonomous vehicles that will later travel on the road segment to update an existing autonomous vehicle road navigation model provided in the navigation systems of the vehicles. The autonomous vehicle road navigation model may be used when an autonomous vehicle autonomously navigates along the common road segment 1200.

[0273] As described above, the autonomous vehicle road navigation model can be included in a sparse map (e.g., the sparse map 800 shown in FIG. 8). The sparse map 800 can include a sparse record of road geometry and / or data related to landmarks along the road, and can provide sufficient information to guide the autonomous navigation of the autonomous vehicle without requiring excessive data storage. In some embodiments, the autonomous vehicle road navigation model can be stored separately from the sparse map 800 and can use map data from the sparse map 800 when the model is executed for navigation. In some embodiments, the autonomous vehicle road navigation model can use the map data included in the sparse map 800 to determine a target trajectory along the road segment 1200 to guide the autonomous navigation of the autonomous vehicles 1205, 1210, 1215, 1220, and 1225 or other vehicles traveling along the road segment 1200 later. For example, when the autonomous vehicle road navigation model is executed by a processor included in the navigation system of vehicle 1205, the model can cause the processor to compare the trajectory determined based on the navigation information received from vehicle 1205 with a predetermined trajectory included in the sparse map 800 to verify and / or correct the current driving course of vehicle 1205.

[0274] In an autonomous vehicle road navigation model, the geometry of a road feature or a target trajectory can be encoded by a curve in three-dimensional space. In one embodiment, the curve can be a three-dimensional spline that includes one or more connected three-dimensional polynomials. As will be understood by those skilled in the art, a spline can be a numerical function that is defined piecewise by a series of polynomials for fitting data. A spline for fitting three-dimensional geometry data of a road can include a linear spline (first order), a quadratic spline (second order), a cubic spline (third order), or other splines (other orders), or combinations thereof. A spline can include one or more three-dimensional polynomials of different orders that connect (e.g., fit) data points of the three-dimensional geometry data of the road. In some embodiments, the autonomous vehicle road navigation model can include a three-dimensional spline corresponding to a common road segment (e.g., road segment 1200) or a target trajectory along a lane of road segment 1200.

[0275] As described above, the autonomous vehicle road navigation model included in the sparse map may include other information, such as the identification of 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 on each of vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, camera 122 may capture an image of the landmark. A processor (e.g., processor 180, 190, or processing unit 110) provided on vehicle 1205 may process the image of the landmark to extract identification information of the landmark. Instead of the actual image of the landmark, the landmark identification information may be stored in sparse map 800. The landmark identification information may require much less storage space than the actual image. Other sensors or systems (e.g., the GPS system) may also provide specific identification information of the landmark (e.g., the position of the landmark). The landmark may include at least one of a traffic sign, an arrow mark, a lane mark, a broken line lane mark, a traffic signal, a stop line, a direction sign (e.g., a highway exit sign with an arrow indicating direction, a highway sign with an arrow pointing to a different direction or location), a landmark beacon, or a street light pole. The landmark beacon refers to a device (e.g., an RFID device) installed along a road segment that transmits or reflects a signal to a receiver installed on a vehicle, so that when the vehicle passes by the device, the beacon received by the vehicle and the position of the device (e.g., determined from the GPS position of the device) may be used as a landmark included in the autonomous vehicle road navigation model and / or sparse map 800.

[0276] The identification of at least one landmark may include the position of the at least one landmark. The position of the landmark can be determined based on position measurements performed using sensor systems associated with a plurality of vehicles 1205, 1210, 1215, 1220, and 1225 (e.g., a global positioning system, an inertial-based positioning system, a landmark beacon, etc.). In some embodiments, the position of the landmark can be determined by averaging position measurements detected, collected, or received by sensor systems on different vehicles 1205, 1210, 1215, 1220, and 1225 over multiple runs. For example, vehicles 1205, 1210, 1215, 1220, and 1225 may transmit position measurement data to server 1230, and server 1230 may average the position measurements and use the average position measurement as the position of the landmark. The position of the landmark can be continuously refined by measurements received from vehicles in later runs.

[0277] Landmark identification may include the size of the landmark. A processor provided in a vehicle (e.g., 1205) may estimate the physical size of a landmark based on the analysis of an image. The server 1230 may receive a plurality of estimated values of the physical size of the same landmark from different vehicles via different runs. The server 1230 may average the different estimated values to arrive at the physical size of the landmark and store the size of the landmark in the road model. The estimated value of the physical size 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 current speed of the vehicle and the scale of magnification based on the position of the landmark appearing in the image with respect to the zoom focus of the camera. For example, the distance to the landmark may be estimated by Z = V * dt * R / D, where V is the speed of the vehicle, R is the distance in the image from the landmark to the zoom focus at time t1, D is the change in the distance of the landmark in the image from t1 to t2, and dt represents (t2 - t1). For example, the distance to the landmark may be estimated by Z = V * dt * R / D, where V is the speed of the vehicle, R is the distance in the image between the landmark and the zoom focus, dt is the time interval, and D is the image displacement of the landmark along the epipolar line. Other equations equivalent to the above equations such as Z = V * ω / Δω may be used to estimate the distance to the landmark. Here, V is the speed of the vehicle, ω is the length of the image (such as the object width), and Δω is the change per unit time of the length of the image.

[0278] When the physical size of the landmark is known, the distance to the landmark can also be determined based on the following equation. 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 passes through the image. From the above equation, the change in the distance Z is ΔZ = f * W * Δω / ω 2 + f * ΔW / ω can be calculated, where ΔW decays to zero by averaging, and Δω is the number of pixels representing the accuracy of the bounding box in the image. The value for estimating the physical size of the landmark can be calculated by averaging multiple observations on the server side. The error resulting from the distance estimation can be very small. There are two sources of error that may occur when using the above equation, namely, ΔW and Δω. The contribution to the distance error is ΔZ = f * W * Δω / ω 2It is given by f*ΔW / ω. However, ΔW decays to zero by averaging. Therefore, ΔZ is determined by Δω (e.g., the inaccuracy of the bounding box of the image).

[0279] For a landmark of unknown dimension, the distance to the landmark can be estimated by tracking feature points on the landmark between consecutive frames. For example, certain features displayed on a speed limit sign can be tracked between two or more image frames. Based on these tracked features, a distance distribution for each feature point can be generated. The distance estimate can be extracted from the distance distribution. For example, the most frequent distance appearing in the distance distribution can be used as the distance estimate. As another example, the average of the distance distribution can be used as the distance estimate.

[0280] FIG. 13 shows an exemplary autonomous vehicle road navigation model represented by a plurality of 3D splines 1301, 1302, and 1303. The curves 1301, 1302, and 1303 shown in FIG. 13 are for illustrative purposes only. Each spline can include one or more 3D polynomials connecting a plurality of data points 1310. Each polynomial can be a first-degree polynomial, a second-degree polynomial, a third-degree polynomial, or any suitable combination of polynomials having different degrees. Each data point 1310 can be associated with navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, each data point 1310 can be associated with data related to a landmark (e.g., size, location, and identification information of the landmark) and / or a road signature profile (e.g., road geometry, road roughness profile, road curvature profile, road width profile). In some embodiments, some of the data points 1310 can be associated with data related to landmark-related data, and other data points can be associated with data related to road signature profile-related data.

[0281] FIG. 14 shows raw position data 1410 (e.g., GPS data) received from five separate runs. A run can be separate from another run if different vehicles cross simultaneously, if the same vehicle crosses at separate times, or if different vehicles cross at separate times. To account for errors in the position data 1410 and different positions of vehicles within the same lane (e.g., one vehicle can be closer to the left side of the lane than another vehicle), the remote server 1230 uses one or more statistical techniques to generate a map skeleton 1420 and can determine whether changes in the raw position data 1410 represent actual deviations or statistical errors. Each path within the skeleton 1420 can be re-linked to the raw data 1410 that formed that path. For example, the path between A and B within the skeleton 1420 is linked to the raw data 1410 from runs 2, 3, 4, and 5, but not from run 1. The skeleton 1420 may not be detailed enough for use in navigating a vehicle (e.g., unlike the splines described above, to combine runs from multiple lanes on the same road), but can provide useful topological information and can be used to define intersections.

[0282] FIG. 15 shows an example of generating additional details for a sparse map within a section of a map skeleton (e.g., section A to B within skeleton 1420). As shown in FIG. 15, data (e.g., self-motion data, road mark data, etc.) can be shown as a function of a position S (or S1 or S2) along a drive. Server 1230 can identify landmarks of a 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 can lead to the identification of landmarks 1511, 1513, and 1515. However, those skilled in the art will recognize that other matching algorithms can be used. For example, optimization of probability can be used instead of, or in combination with, a unique match. Server 1230 can align the drives vertically and align the matched landmarks. For example, server 1230 can select one drive (e.g., drive 1520) as a reference drive and then shift and / or elastically stretch the other drive (e.g., drive 1510) for alignment.

[0283] FIG. 16 shows an example of aligned landmark data for use in a sparse map. In the example of FIG. 16, landmark 1610 includes a road sign. The example of FIG. 16 further shows data from a plurality of drives 1601, 1603, 1605, 1607, 1609, 1611, and 1613. In the example of FIG. 16, the data from drive 1613 is composed of "ghost" landmarks, and since none of drives 1601, 1603, 1605, 1607, 1609, and 1611 include the identification of a landmark in the vicinity of an identified landmark within drive 1613, server 1230 can identify the landmark as a "ghost". Thus, server 1230 can accept a potential landmark if the ratio of images in which a landmark appears to images in which the landmark does not appear exceeds a threshold, and / or can reject a potential landmark if the ratio of images in which a landmark does not appear to images in which the landmark appears exceeds a threshold.

[0284] Figure 17 shows a system 1700 for generating driving data that can be used to crowdsource sparse maps. As shown in Figure 17, the system 1700 may include a camera 1701 and a location device 1703 (e.g., a GPS locator). The camera 1701 and the location device 1703 may be mounted on a vehicle (e.g., one of vehicles 1205, 1210, 1215, 1220, and 1225). The camera 1701 may generate multiple types of data, such as ego-motion data, traffic sign data, road data, etc. The camera data and the location data may be divided into driving segments 1705. For example, each of the driving segments 1705 may have camera data and location data for a drive of less than 1 km.

[0285] In some embodiments, the system 1700 may remove redundancy in the driving segments 1705. For example, if a landmark appears in multiple images from the camera 1701, the system 1700 may remove redundant data such that the driving segment 1705 includes only one copy of the location of the landmark and the metadata associated with the landmark. As a further example, if a lane mark appears in multiple images from the camera 1701, the system 1700 may remove redundant data such that the driving segment 1705 includes only one copy of the location of the lane mark and the metadata associated with the lane mark.

[0286] The system 1700 also includes a server (e.g., server 1230). The server 1230 may receive the driving segments 1705 from the vehicle and recombine the driving segments 1705 into a single drive 1707. Such an arrangement may reduce the bandwidth requirements when transferring data between the vehicle and the server, and the server may also store data related to the entire drive.

[0287] FIG. 18 shows the system 1700 of FIG. 17 further configured to crowdsource a sparse map. As in FIG. 17, the system 1700 includes, for example, a vehicle 1810 that captures driving data using, for example, a camera (e.g., generating self-motion data, traffic sign data, road data, etc.) and a positioning device (e.g., a GPS locator). As in FIG. 17, the vehicle 1810 divides the collected data into driving segments (shown as "DS1 1", "DS2 1", "DSN 1" in FIG. 18). The server 1230 then receives the driving segments and reconstructs a drive (shown as "Drive 1" in FIG. 18) from the received segments.

[0288] As further shown in FIG. 18, the system 1700 also receives data from additional vehicles. For example, vehicle 1820 also captures driving data using, for example, a camera (e.g., generating self-motion data, traffic sign data, road data, etc.) and a positioning device (e.g., a GPS locator). Similar to vehicle 1810, vehicle 1820 divides the data collected into driving segments (shown as "DS1 2", "DS2 2", "DSN 2" in FIG. 18). The server 1230 then receives the driving segments and reconstructs a drive (shown as "Drive 2" in FIG. 18) from the received segments. Any number of additional vehicles can be used. For example, FIG. 18 also includes "Vehicle N" that captures driving data, divides it into driving segments (shown as "DS1 N", "DS2 N", "DSN N" in FIG. 18), and transmits it to the server 1230 for reconstruction into a drive (shown as "Drive N" in FIG. 18).

[0289] As shown in FIG. 18, the server 1230 can construct a sparse map (shown as "Map") using the reconstructed drives (e.g., "Drive 1", "Drive 2", and "Drive N") collected from a plurality of vehicles (e.g., "Vehicle 1" (also denoted as vehicle 1810), "Vehicle 2" (also denoted as vehicle 1820), and "Vehicle N").

[0290] FIG. 19 is a flowchart illustrating an exemplary process 1900 for generating a sparse map for autonomous vehicle navigation along a road segment. The process 1900 can be executed by one or more processing devices included in the server 1230.

[0291] The process 1900 can include receiving a plurality of images acquired when one or more vehicles cross a road segment (step 1905). The server 1230 can receive images from cameras included in one or more of the vehicles 1205, 1210, 1215, 1220, and 1225. For example, the camera 122 can capture one or more images of the environment surrounding the vehicle 1205 as the vehicle 1205 travels along the road segment 1200. In some embodiments, the server 1230 can receive pre - removed image data with redundancy removed by a processor on the vehicle 1205, as discussed above with respect to FIG. 17.

[0292] The process 1900 can further include identifying at least one line representation of a road surface feature extending along the road segment based on the plurality of images (step 1910). Each line representation can represent a path along the road segment that substantially corresponds to the road surface feature. For example, the server 1230 can analyze the environmental images received from the camera 122 to identify road edges or lane markings and determine a trajectory of travel along the road segment 1200 associated with the road edges or lane markings. In some embodiments, the trajectory (or line representation) can include a spline, a polynomial representation, or a curve. The server 1230 can determine the trajectory of travel of the vehicle 1205 based on the self - motion of the camera (e.g., 3 - dimensional translational motion and / or 3 - dimensional rotational motion) received in step 1905.

[0293] Process 1900 may also include identifying a plurality of landmarks associated with a road segment based on a plurality of images (step 1910). For example, server 1230 may analyze the 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 the analysis of a plurality of images acquired when one or more vehicles cross the road segment. To enable crowdsourcing, the analysis may include rules regarding the acceptance and rejection of potential landmarks associated with the road segment. For example, the analysis may include accepting 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, and / or 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.

[0294] Process 1900 may include other operations or steps performed by server 1230. For example, the navigation information may include a target trajectory for a vehicle to travel along the road segment, and process 1900 may include clustering vehicle trajectories associated with a plurality of vehicles traveling on the road segment and determining a target trajectory based on the clustered vehicle trajectories by server 1230, as discussed in more detail below. Clustering the vehicle trajectories may include clustering a plurality of trajectories associated with vehicles traveling on the road segment into a plurality of clusters by server 1230 based on at least one of the absolute direction of travel of the vehicle or the lane assignment of the vehicle. Generating the target trajectory may include averaging the clustered trajectories by server 1230. As a further example, process 1900 may include aligning the data received in step 1905. Other processes or steps performed by server 1230, as described above, may also be included in process 1900.

[0295] The disclosed systems and methods may include other features. For example, the disclosed systems may use local coordinates instead of global coordinates. In the case of autonomous driving, some systems may display data in global coordinates. For example, the longitude and latitude coordinates of the ground surface may be used. To use a map for steering, the host vehicle may determine its position and orientation relative to the map. It would seem natural to use an in-vehicle GPS device to place the vehicle on the map and find the rotational transformation between the body's reference frame and the world's reference frame (e.g., north, east, and down). Once the body's reference frame is aligned with the map's reference frame, the desired route may be represented in the body's reference frame, and steering commands may be calculated or generated.

[0296] The disclosed systems and methods may enable autonomous vehicle navigation (e.g., steering control) in a low footprint model that can be collected by the autonomous vehicle itself without the aid of expensive surveying equipment. To support autonomous navigation (e.g., a steering application), the road model may include a sparse map with the geometry of the road, its lane structure, and landmarks that can be used to determine the location or position of the vehicle along the tracks included in the model. As discussed above, the generation of the sparse map may be performed by a remote server that communicates with the vehicle driving on the road and receives data from the vehicle. The data may include the detected data, the tracks reconstructed based on the detected data, and / or the recommended tracks that may represent the reconstructed tracks to be corrected. As discussed below, the server may transmit the model to vehicles that will later drive on the road or other vehicles to assist in autonomous navigation.

[0297] FIG. 20 shows a block diagram of server 1230. Server 1230 may include a communication unit 2005 that includes both hardware components (e.g., communication control circuits, switches, and antennas) and software components (e.g., communication protocols, computer code). For example, communication unit 2005 may include at least one network interface. Server 1230 may communicate with vehicles 1205, 1210, 1215, 1220, and 1225 through communication unit 2005. For example, server 1230 may receive navigation information transmitted from vehicles 1205, 1210, 1215, 1220, and 1225 through communication unit 2005. Server 1230 may distribute an autonomous vehicle road navigation model to one or more autonomous vehicles through communication unit 2005.

[0298] Server 1230 may include at least one non-transitory storage medium 2010 such as a hard drive, compact disk, tape, etc. Storage device 1410 may be configured to store navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225 and / or data such as an autonomous vehicle road navigation model generated by server 1230 based on the navigation information. Storage device 2010 may be configured to store other information such as a sparse map (e.g., sparse map 800 discussed above with respect to FIG. 8).

[0299] In addition to or instead of storage device 2010, server 1230 may include a memory 2015. Memory 2015 may be the same as or different from memory 140 or 150. Memory 2015 may be a non-transitory memory such as flash memory, random access memory, etc. Memory 2015 may be configured to store computer code or instructions executable by a processor (e.g., processor 2020), map data (e.g., data of sparse map 800), an autonomous vehicle road navigation model, and / or data such as navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225.

[0300] Server 1230 may include at least one processing device 2020 configured to execute computer code or instructions stored in memory 2015 to perform various functions. For example, processing device 2020 may analyze navigation information received from vehicles 1205, 1210, 1215, 1220, and 1225 and generate an autonomous vehicle road navigation model based on the analysis. Processing device 2020 may control communication unit 1405 to distribute the autonomous vehicle road navigation model to one or more autonomous vehicles (e.g., one or more of vehicles 1205, 1210, 1215, 1220, and 1225, or any vehicle that will later travel on road segment 1200). Processing device 2020 may be similar to or different from processors 180, 190, or processing unit 110.

[0301] FIG. 21 shows a block diagram of memory 2015 that may store computer code or instructions for performing one or more operations for generating a road navigation model for use in autonomous vehicle navigation. As shown in FIG. 21, memory 2015 may store one or more modules for performing operations for processing vehicle navigation information. For example, memory 2015 may include a model generation module 2105 and a model distribution module 2110. Processor 2020 may execute instructions stored in either of modules 2105 and 2110 included in memory 2015.

[0302] When executed by the processor 2020, the model generation module 2105 can store instructions that can generate at least a part of an autonomous vehicle 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, in generating the autonomous vehicle road navigation model, the processor 2020 can cluster vehicle trajectories along the common road segment 1200 into different clusters. The processor 2020 can determine a target trajectory along the common road segment 1200 based on the clustered vehicle trajectories for each of the different clusters. Such an operation can include finding the average or mean trajectory of the clustered vehicle trajectories in each cluster (e.g., by averaging the data representing the clustered vehicle trajectories). In some embodiments, the target trajectory can be associated with a single lane of the common road segment 1200.

[0303] The road model and / or the sparse map can store trajectories associated with the road segment. These trajectories can be called target trajectories and are provided to the autonomous vehicle for autonomous navigation. The target trajectories can be received from multiple vehicles and can be generated based on the actual trajectories or the recommended trajectories (actual trajectories with some modifications) received from multiple vehicles. The target trajectories included in the road model or the sparse map can be continuously updated (e.g., averaged) with new trajectories received from other vehicles.

[0304] Vehicles traveling on a road segment can collect data with various sensors. The data can include landmarks, road signature profiles, vehicle movement (e.g., accelerometer data, speed data), vehicle position (e.g., GPS data), and can reconstruct the actual trajectory itself or transmit the data to a server, where the server can reconstruct the actual trajectory of the vehicle. In some embodiments, the vehicle can transmit data related to a trajectory (e.g., a curve in any reference frame), landmark data, and lane assignment along the driving route to server 1230. Various vehicles traveling along the same road segment in multiple trips can have different trajectories. Server 1230 can identify the route or trajectories associated with each lane from the trajectories received from the vehicles through a clustering process.

[0305] FIG. 22 shows a process of clustering vehicle trajectories associated with vehicles 1205, 1210, 1215, 1220, and 1225 to determine a target trajectory for a common road segment (e.g., road segment 1200). The target trajectory or trajectories determined from the clustering process can be included in an autonomous vehicle road navigation model or a sparse map 800. In some embodiments, vehicles 1205, 1210, 1215, 1220, and 1225 traveling along road segment 1200 can transmit a plurality of trajectories 2200 to server 1230. In some embodiments, server 1230 can generate a trajectory based on landmarks, road geometry, and vehicle movement information received from vehicles 1205, 1210, 1215, 1220, and 1225. To generate an autonomous vehicle road navigation model, server 1230 can cluster vehicle trajectories 1600 into a plurality of clusters 2205, 2210, 2215, 2220, 2225, and 2230 as shown in FIG. 22.

[0306] Clustering can be performed using various criteria. In some embodiments, all the runs within a cluster can be similar with respect to the absolute direction of travel along road segment 1200. The absolute direction of travel can be obtained from the GPS signals received by vehicles 1205, 1210, 1215, 1220, and 1225. In some embodiments, dead reckoning can be used to obtain the absolute direction of travel. As those skilled in the art will understand, dead reckoning can be used to determine the current position, and thus the direction of travel of vehicles 1205, 1210, 1215, 1220, and 1225, using previously determined positions, estimated speeds, etc. The trajectories clustered by the absolute direction of travel can be useful for identifying the route along the road.

[0307] In some embodiments, all the runs within a cluster can be similar with respect to the lane assignment (e.g., the same lane before and after an intersection) along the run of road segment 1200. The trajectories clustered by the lane assignment can be useful for identifying the lane along the road. In some embodiments, both criteria (e.g., absolute direction of travel and lane assignment) can be used for clustering.

[0308] For each of clusters 2205, 2210, 2215, 2220, 2225, and 2230, the trajectories can be averaged to obtain a target trajectory associated with a particular cluster. For example, the trajectories from multiple runs associated with the same lane cluster can be averaged. The average trajectory can be the target trajectory associated with a particular lane. To average the cluster of trajectories, server 1230 can select a reference frame for any trajectory C0. For all other trajectories (C1,..., Cn), server 1230 can find a rigid body transformation that maps Ci to C0, where i = 1, 2,..., n, and n is a positive integer corresponding to the total number of trajectories included in the cluster. Server 1230 can calculate the average curve or trajectory in the C0 reference frame.

[0309] In some embodiments, the landmarks may define an arc length that is consistent across different runs, and the arc length may be used to align the trajectory with the lane. In some embodiments, lane markings before and after an intersection may be used to align the trajectory with the lane.

[0310] To assemble 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 in the same reference frame. Adjacent lanes may be aligned as if they were the same lane and later shifted horizontally.

[0311] Landmarks recognized along a road segment may be first mapped to a common reference frame at the lane level and then at the intersection level. For example, the same landmark may be recognized multiple times by multiple vehicles in multiple runs. Data regarding the same landmark received in different runs may be slightly different. Such data may be averaged and mapped to the same reference frame, such as the C0 reference frame. Additionally or alternatively, the variance of data for the same landmark received in multiple runs may be calculated.

[0312] In some embodiments, each lane of road segment 120 may be associated with a target trajectory and specific landmarks. The target trajectory or such multiple target trajectories may be included in an autonomous vehicle road navigation model, and the autonomous vehicle road navigation model may later be used by other autonomous vehicles traveling along the same road segment 1200. While vehicles 1205, 1210, 1215, 1220, and 1225 are traveling along road segment 1200, the landmarks identified by those vehicles may be recorded in association with the target trajectory. The data of the target trajectory and the landmarks may be continuously or periodically updated with new data received from other vehicles in subsequent runs.

[0313] For the localization of autonomous vehicles, the disclosed systems and methods may use an extended Kalman filter. The position of the vehicle may be determined based on 3D position data and / or 3D orientation data, and a prediction of a future position of the vehicle ahead of the current position by integration of self-motion. The position of the vehicle may be corrected or adjusted by an image observation of a landmark. For example, if the vehicle detects a landmark in an image captured by a camera, the landmark may be compared with known landmarks stored in a road model or a sparse map 800. The known landmarks may have known positions (e.g., GPS data) along a target trajectory stored in the road model and / or the 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 position of the vehicle along the target trajectory may be adjusted based on the distance to the landmark and the known position of the landmark (stored in the road model or the sparse map 800). The position / location data of the landmarks (e.g., average values from multiple runs) stored in the road model and / or the sparse map 800 may be estimated to be accurate.

[0314] In some embodiments, the disclosed system may form a closed-loop subsystem in which the estimation of the position of the vehicle in 6 degrees of freedom (e.g., 3D position data and 3D orientation data) may be used to navigate (e.g., steer the wheels) the autonomous vehicle to reach a desired point (e.g., 1.3 seconds ahead of a stored point). Next, the data measured from the steering and the actual navigation may be used to estimate the 6-degree-of-freedom position.

[0315] In some embodiments, poles along the road, such as street light poles and power line poles or cable line poles, may be used as landmarks for vehicle localization. Other landmarks, such as traffic signs, traffic lights, arrows on the road, stop lines, and static features or signatures of objects along road segments, may also be used as landmarks for vehicle localization. When a pole is used for localization, since the bottom of the pole may be blocked and not on the road plane, the x-observation of the pole (i.e., the viewing angle from the vehicle) rather than the y-observation (i.e., the distance to the pole) may be used.

[0316] FIG. 23 shows a navigation system for a vehicle that can be used for autonomous navigation using a sparsely mapped cloud-sourced map. For purposes of explanation, the vehicle is referred to as vehicle 1205. The vehicle shown in FIG. 23 can be any of the other vehicles disclosed herein, including, for example, vehicles 1210, 1215, 1220, and 1225, as well as vehicle 200 shown in other embodiments. As shown in FIG. 12, vehicle 1205 can communicate with server 1230. Vehicle 1205 can include an image capture device 122 (e.g., camera 122). Vehicle 1205 can include a navigation system 2300 configured to provide navigation guidance for vehicle 1205 to travel on a road (e.g., road segment 1200). Vehicle 1205 can also include other sensors such as a speed sensor 2320 and an accelerometer 2325. The speed sensor 2320 can be configured to detect the speed of vehicle 1205. The accelerometer 2325 can be configured to detect the acceleration or deceleration of vehicle 1205. The vehicle 1205 shown in FIG. 23 can be an autonomous vehicle, and the navigation system 2300 can be used to provide navigation guidance for autonomous driving. Alternatively, vehicle 1205 can be a non-autonomous human-controlled vehicle, and the navigation system 2300 can still be used to provide navigation guidance.

[0317] The navigation system 2300 may include a communication unit 2305 configured to communicate with the server 1230 through the 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 the sparse map 800 (which may be stored in a storage device mounted on the vehicle 1205 and / or received from the server 1230), the geometry of the road sensed by the road profile sensor 2330, images captured by the camera 122, and / or an autonomous vehicle road navigation model 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 the roughness of the road surface, road width, road height, road curvature, etc. For example, the road profile sensor 2330 may include a device that measures the movement of the suspension of the vehicle 2305 to derive a road roughness profile. In some embodiments, the road profile sensor 2330 includes a radar sensor and may measure the distance from the vehicle 1205 to the road side (e.g., a road side barrier), thereby measuring the width of the road. In some embodiments, the road profile sensor 2330 may include a device configured to measure the elevation of the road up and down. In some embodiments, the road profile sensor 2330 may include a device configured to measure the road curvature. For example, a camera (e.g., the camera 122 or another camera) may be used to capture an image of the road indicating the road curvature. The vehicle 1205 may use such an image to detect the road curvature.

[0318] At least one processor 2315 can be programmed to receive from camera 122 at least one environmental image associated with vehicle 1205. The at least one processor 2315 can analyze the at least one environmental image to determine navigation information associated with vehicle 1205. The navigation information can include a trajectory related to the travel of vehicle 1205 along road segment 1200. The at least one processor 2315 can determine the trajectory based on the movement of camera 122 (and thus the vehicle), such as 3D translational movement and 3D rotational movement. In some embodiments, the at least one processor 2315 can determine the translational and rotational movement of camera 122 based on the analysis of a plurality of images acquired by camera 122. In some embodiments, the navigation information can include lane assignment information (e.g., whether vehicle 1205 is traveling in its lane along road segment 1200). The navigation information transmitted from vehicle 1205 to server 1230 can be used by server 1230 to generate and / or update an autonomous vehicle road navigation model, and the autonomous vehicle road navigation model can be transmitted from server 1230 to vehicle 1205 to provide autonomous navigation guidance for vehicle 1205.

[0319] At least one processor 2315 can also be programmed to transmit navigation information from vehicle 1205 to server 1230. In some embodiments, the navigation information can be transmitted to server 1230 along with road information. The road location information can include at least one of GPS signals received by GPS unit 2310, landmark information, road geometry, lane information, etc. At least one processor 2315 can receive from server 1230 an autonomous vehicle road navigation model or a part of the model. The autonomous vehicle road navigation model received from server 1230 can include at least one update based on the navigation information transmitted from vehicle 1205 to server 1230. The part of the model transmitted from server 1230 to vehicle 1205 can include the updated part of the model. At least one processor 2315 can cause at least one navigation operation (e.g., steering such as making a turn, applying brakes, accelerating, overtaking another vehicle, etc.) by vehicle 1205 based on the received autonomous vehicle road navigation model or the updated part of the model.

[0320] At least one processor 2315 can be configured to communicate with various sensors and components included in vehicle 1205, including communication unit 1705, GPS unit 2315, camera 122, speed sensor 2320, accelerometer 2325, and road profile sensor 2330. At least one processor 2315 can collect information or data from the various sensors and components and transmit the information or data to server 1230 through communication unit 2305. Alternatively or additionally, the various sensors or components of vehicle 1205 can also communicate with server 1230 and transmit the data or information collected by the sensors or components to server 1230.

[0321] In some embodiments, at least one of vehicles 1205, 1210, 1215, 1220, and 1225 can communicate with each other and share navigation information with each other so that, for example, based on information shared by other vehicles, an autonomous vehicle road navigation model can be generated using crowdsourcing. In some embodiments, vehicles 1205, 1210, 1215, 1220, and 1225 can share navigation information with each other, and each vehicle can update its own autonomous vehicle road navigation model provided to the vehicle. In some embodiments, at least one of vehicles 1205, 1210, 1215, 1220, and 1225 (e.g., vehicle 1205) can function as a hub vehicle. At least one processor 2315 of the hub vehicle (e.g., vehicle 1205) can execute some or all of the functions executed by server 1230. For example, at least one processor 2315 of the hub vehicle can communicate with other vehicles and receive navigation information from other vehicles. At least one processor 2315 of the hub vehicle can generate an autonomous vehicle road navigation model or an update to the model based on the shared information received from other vehicles. At least one processor 2315 of the hub vehicle can send the autonomous vehicle road navigation model or the update to the model to other vehicles to provide autonomous navigation guidance.

[0322] Lane Mark Mapping and Navigation Based on Mapped Lane Marks

[0323] As described above, the autonomous vehicle's road navigation model and / or sparse map 800 may include a plurality of mapped lane marks associated with road segments. As discussed in more detail below, these mapped lane marks can be used when the autonomous vehicle is navigating. For example, in some embodiments, the mapped lane marks can be used to determine the lateral position and / or orientation relative to a planned trajectory. Using this position information, the autonomous vehicle's travel direction can be adjusted to align with the direction of the target trajectory at the determined position.

[0324] Vehicle 200 may be configured to detect lane marks in a given road segment. A road segment may include any mark on the road for guiding the traffic volume of vehicles on the road. For example, a lane mark may be a solid or dashed line indicating the end of a driving lane. A lane mark may also include, for example, double lines such as double solid lines, double dashed lines, or a combination of solid and dashed lines indicating whether passing in an adjacent lane is permitted. A lane mark may also include, for example, highway entrance and exit marks indicating a deceleration lane of an exit ramp, or a dotted line indicating that a lane is for a direction change only or that a lane has ended. The mark may further indicate a work area, a temporary lane shift, a driving route through an intersection, a median strip, a dedicated lane (e.g., a bicycle lane, an HOV lane, etc.), or other various miscellaneous marks (e.g., a crosswalk, a speed bump, a railroad crossing, a stop line, etc.).

[0325] Vehicle 200 can capture images of surrounding lane marks using cameras such as image capture devices 122 and 124 included in image acquisition unit 120. Vehicle 200 can analyze the images to detect point positions associated with the lane marks based on features identified within one or more of the captured images. These point positions can be uploaded to a server to represent the lane marks of sparse map 800. Depending on the position and field of view of the cameras, lane marks on both sides of the vehicle can be detected simultaneously from a single image. In other embodiments, different cameras can be used to capture images at multiple sides of the vehicle. Instead of uploading actual images of the lane marks, the marks can be stored in sparse map 800 as splines or a series of points, thus reducing the size of sparse map 800 and / or the data that must be remotely uploaded by the vehicle.

[0326] Figures 24A - 24D show exemplary point positions that can be detected by vehicle 200 to represent a specific lane mark. Similar to the above - mentioned landmarks, vehicle 200 can use various image - recognition algorithms or software to identify point positions within the captured images. For example, vehicle 200 can recognize a series of end points, corner points, or other various point positions associated with a specific lane mark. Figure 24A shows a continuous lane mark 2410 that can be detected by vehicle 200. Lane mark 2410 can represent the outer edge of a road represented by a continuous white line. As shown in Figure 24A, vehicle 200 can be configured to detect a plurality of end - point positions 2411 along the lane mark. The position points 2411 can be collected to represent the lane mark at any interval sufficient to create a mapped lane mark within a sparse map. For example, the lane mark can be represented by one point per 1 meter of the detected end, one point per 5 meters of the detected end, or other suitable intervals. In some embodiments, the interval can be determined by other factors rather than a set interval, such as, for example, based on the point with the highest reliability ranking of the position of the points detected by vehicle 200. Figure 24A shows end - point positions on the inner edge of lane mark 2410, but the points can be collected on the outer edge of the line or along both edges. Further, although a single line is shown in Figure 24A, similar end points can be detected for a double solid line. For example, points 2411 can be detected along one or both ends of the solid line.

[0327] Vehicle 200 may also represent different lane marks depending on the type or shape of the lane mark. FIG. 24B shows an exemplary dashed lane mark 2420 that may be detected by vehicle 200. Instead of identifying endpoints as in FIG. 24A, the vehicle may detect a series of corner points 2421 that represent the corners of the lane dashes to define the complete boundary of the dashed line. FIG. 24B shows that each corner of a given dashed mark is located, but vehicle 200 may detect or upload a subset of the points shown in the figure. For example, vehicle 200 may detect the front end or front corner of a given dashed mark, or the two corner points closest to the inside of the lane. Further, not all dashed marks may be captured. For example, vehicle 200 may capture and / or record points representing samples (e.g., every other, every third, every fifth, etc.) of the dashed mark, or points representing the dashed mark at predefined intervals (e.g., every meter, every five meters, every ten meters, etc.). Corner points may also be detected for similar lane marks such as marks indicating that a lane is for an exit ramp, marks indicating that a particular lane is about to end, or other various lane marks that may have detectable corner points. Corner points may also be detected for lane marks composed of double dashed lines or combinations of solid and dashed lines.

[0328] In some embodiments, the points uploaded to the server to generate the mapped lane marks may represent other points than the detected end points or corner points. FIG. 24C shows a series of points that may represent the center line of a given lane mark. For example, the continuous lane 2410 may be represented by center line points 2441 along the center line 2440 of the lane mark. In some embodiments, the vehicle 200 may be configured to detect these center points using various image recognition techniques such as convolutional neural networks (CNNs), scale-invariant feature transforms (SIFTs), histograms of oriented gradients (HOG) features, or other techniques. Alternatively, the vehicle 200 may detect other points such as the end point 2411 shown in FIG. 24A and calculate the center line point 2441, for example, by detecting points along each end and determining the midpoint between the end points. Similarly, the dashed lane mark 2420 may be represented by center line points 2451 along the center line 2450 of the lane mark. The center line points may be located at the ends of the dashed line, or at various other positions along the center line, as shown in FIG. 24C. For example, each dashed line may be represented by a single point at the geometric center of the dashed line. The points may also be spaced at a predetermined interval (e.g., every 1 meter, every 5 meters, every 10 meters, etc.) along the center line. The center line point 2451 may be directly detected by the vehicle 200 or calculated based on other detected reference points such as the corner point 2421 as shown in FIG. 24B. The center line may also be used to represent other lane mark types such as double lines using techniques similar to those described above.

[0329] In some embodiments, vehicle 200 may identify points representing other features, such as vertices between two intersecting lane marks. FIG. 24D shows an exemplary point representing an intersection between two lane marks 2460 and 2465. Vehicle 200 may calculate a vertex 2466 representing the intersection between the two lane marks. For example, one of the lane marks 2460 or 2465 may represent a train intersection area or other intersection area within a road segment. Lane marks 2460 and 2465 are shown as intersecting perpendicularly to each other, but various other configurations may be detected. For example, lane marks 2460 and 2465 may intersect at other angles, or one or both of the lane marks may end at vertex 2466. Similar techniques may be applied to intersections between dashed lines or other lane mark types. In addition to vertex 2466, various other points 2467 may also be detected, providing additional information about the orientation of lane marks 2460 and 2465.

[0330] Vehicle 200 can associate real-world coordinates with each detected point of the lane mark. For example, a position identifier including the coordinates of each point can be generated and uploaded to the server for mapping the lane mark. The position identifier can further include other identification information regarding the point, including whether the point represents a corner point, an end point, a center point, etc. Accordingly, vehicle 200 can be configured to determine the real-world position of each point based on the analysis of the image. For example, vehicle 200 can detect other features in the image such as the various landmarks described above to identify the real-world position of the lane mark. This can include determining the position of the lane mark in the image relative to the detected landmark, or determining the position of the vehicle based on the detected landmark and then determining the distance from the vehicle (or the target trajectory of the vehicle) to the lane mark. If no landmark is available, the position of the lane mark point can be determined based on the position of the vehicle determined based on dead reckoning. The real-world coordinates included in the position identifier can be represented as absolute coordinates (e.g., latitude / longitude coordinates) or can be related to other features, such as based on the longitudinal position along the target trajectory and the lateral distance from the target trajectory. The position identifier can then be uploaded to the server to generate a lane mark mapped in a navigation model (such as sparse map 800). In some embodiments, the server can construct a spline representing the lane mark of the road segment. Alternatively, vehicle 200 can generate the spline and upload it to the server for recording in the navigation model.

[0331] FIG. 24E shows an example of a corresponding road segment's exemplary navigation model or sparse map including the mapped lane mark. The sparse map can include a target trajectory 2475 that the vehicle follows along the road segment. As described above, the target trajectory 2475 can represent the ideal path that the vehicle passes through when traveling on the corresponding road segment or can be located at other places on the road (e.g., the center line of the road, etc.). The target trajectory 2475 can be calculated in various ways as described above, for example, based on the aggregation (e.g., weighted combination) of two or more reconstructed trajectories of vehicles crossing the same road segment.

[0332] In some embodiments, the target trajectory can be generated equally for all vehicle types and all road, vehicle, and / or environmental conditions. However, in other embodiments, various other factors or variables can also be considered when generating the target trajectory. Different target trajectories can be generated for different types of vehicles (e.g., passenger cars, light trucks, and full trailers). For example, for a small passenger car, a target trajectory with a relatively narrow turning radius can be generated compared to a large semi-trailer truck. In some embodiments, road, vehicle, and environmental conditions can also be considered. For example, different target trajectories can be generated for different road conditions (e.g., wet, snowy, icy, dry, etc.), vehicle conditions (e.g., tire conditions or estimated tire conditions, brake conditions or estimated brake conditions, remaining fuel amount, etc.) or environmental factors (e.g., time, visibility, weather, etc.). The target trajectory can also depend on one or more aspects or features of a particular road segment (e.g., speed limit, frequency and size of direction changes, gradient, etc.). In some embodiments, various user settings can be used to determine the target trajectory such as a set driving mode (e.g., desired aggressive driving, economy mode, etc.).

[0333] The sparse map may also include mapped lane marks 2470 and 2480 that represent lane marks along the road segment. The mapped lane marks may be represented by a plurality of position identifiers 2471 and 2481. As described above, the position identifier may include the position in the real-world coordinates of the points associated with the detected lane marks. Similar to the target trajectory of the model, the lane marks also include elevation data and may be represented as a curve in three-dimensional space. For example, the curve may be a spline connecting three-dimensional polynomials of an appropriate degree, or the curve may be calculated based on the position identifiers. The mapped lane marks may also include other information or metadata about the lane marks, such as an identifier of the type of lane mark (e.g., between two lanes having the same travel direction, between two lanes having opposite travel directions, at the end of the road, etc.) and / or other characteristics of the lane mark (e.g., solid line, dashed line, single line, double line, yellow line, white line, etc.). In some embodiments, the mapped lane marks may be continuously updated in the model using, for example, crowdsourcing techniques. The same vehicle may upload the position identifiers during multiple opportunities to travel the same road segment, or the data may be selected from a plurality of vehicles (such as 1205, 1210, 1215, 1220, and 1225, etc.) traveling the road segment at different times. Then, the sparse map 800 can be updated or refined based on subsequent position identifiers received from the vehicle and stored in the system. When the mapped lane marks are updated and refined, the updated road navigation model and / or the sparse map can be distributed to a plurality of autonomous vehicles.

[0334] Generating lane marks mapped within a sparse map may also include detecting and / or reducing errors based on anomalies in the image or the actual lane marks themselves. FIG. 24F shows an exemplary anomaly 2495 associated with the detection of lane mark 2490. Anomaly 2495 may appear in an image captured by vehicle 200, for example, from an object blocking the camera's view of the lane mark, dirt on the lens, etc. In some cases, the anomaly may be due to the lane mark itself, which may be damaged, worn, or partially covered by, for example, dirt, debris, water, snow, or other substances on the road. Anomaly 2495 may cause an incorrect point 2491 to be detected by vehicle 200. Sparse map 800 may provide correctly mapped lane marks and exclude errors. In some embodiments, vehicle 200 may detect incorrect point 2491, for example, by detecting anomaly 2495 in the image or by identifying an error based on lane mark points detected before and after the anomaly. Based on the detection of the anomaly, the vehicle may exclude point 2491 or adjust it to match other detected points. In other embodiments, the error may be corrected by determining that the point is outside the expected threshold range, for example, based on other points uploaded during the same drive or based on an aggregation of data from previous drives along the same road segment, after the point has been uploaded.

[0335] Lane marks mapped to a navigation model and / or a sparse map can also be used for the navigation of an autonomous vehicle across a corresponding road. For example, a vehicle navigating along a target trajectory can periodically use the mapped lane marks in the sparse map to align itself with the target trajectory. As described above, between landmarks, the vehicle can navigate based on dead reckoning where the vehicle uses sensors to determine its self-motion and estimate its position relative to the target trajectory. Errors can accumulate over time, and the accuracy of the vehicle's positioning relative to the target trajectory can gradually decrease. Thus, the vehicle can use lane marks generated in the sparse map 800 (and their known positions) to reduce the errors induced by dead reckoning in positioning. In this way, the identified lane marks included in the sparse map 800 can function as navigation anchors from which the accurate position of the vehicle relative to the target trajectory can be determined.

[0336] FIG. 25A shows an exemplary image 2500 of the surrounding environment of a vehicle that can be used for navigation based on mapped lane marks. The image 2500 can be captured by the vehicle 200 via, for example, the image capture devices 122 and 124 included in the image acquisition unit 120. The image 2500 can include an image of at least one lane mark 2510, as shown in FIG. 25A. The image 2500 can also include one or more landmarks 2521, such as road signs, used for navigation as described above. Although not shown in the captured image 2500, some elements shown in FIG. 25A, such as elements 2511, 2530, and 2520 detected and / or determined by the vehicle 200, are also shown for reference.

[0337] Using the various techniques described above with respect to FIGS. 24A-24D and 24F, the vehicle can analyze Image 2500 to identify lane markings 2510. Various points 2511 corresponding to features of the lane markings within the image can be detected. For example, points 2511 can correspond to the ends of lane markings, the corners of lane markings, the midpoints of lane markings, the vertices between two intersecting lane markings, or other various features or locations. Points 2511 can be detected to correspond to the positions of points stored in a navigation model received from a server. For example, when a sparse map including points representing the centerlines of mapped lane markings is received, points 2511 can also be detected based on the centerlines of lane markings 2510.

[0338] The vehicle can also be represented by element 2520 and determine a longitudinal position disposed along a target trajectory. The longitudinal position 2520 can be determined from Image 2500, for example, by detecting landmarks 2521 within Image 2500 and comparing the measured positions to known landmark positions stored in a road model or sparse map 800. The position of the vehicle along the target trajectory can then be determined based on the distance to the landmark and the known position of the landmark. The longitudinal position 2520 can also be determined from an image other than the one used to determine the position of the lane markings. For example, the longitudinal position 2520 can be determined by detecting landmarks within an image from another camera within image acquisition unit 120 that is taken at the same time as or substantially at the same time as Image 2500. In some cases, the vehicle may not be near a landmark or other reference point for determining the longitudinal position 2520. In such cases, the vehicle can navigate based on dead reckoning and thus use sensors to determine its self-motion and estimate the longitudinal position 2520 relative to the target trajectory. The vehicle can also determine a distance 2530 representing the actual distance between the vehicle observed in the captured image and lane markings 2510. The angle of the camera, the speed of the vehicle, the width of the vehicle, or other various factors can be considered when determining distance 2530.

[0339] FIG. 25B shows a lateral position identification correction of a vehicle based on mapped lane marks in a road navigation model. As described above, vehicle 200 can determine a distance 2530 between vehicle 200 and lane mark 2510 using one or more images captured by vehicle 200. Vehicle 200 can also access a road navigation model such as sparse map 800, which may include mapped lane marks 2550 and target trajectory 2555. The mapped lane marks 2550 can be modeled using the techniques described above, for example, using cloud-sourced position identifiers captured by multiple vehicles. The target trajectory 2555 can also be generated using the various techniques described above. Vehicle 200 can also determine or estimate a longitudinal position 2520 along the target trajectory 2555 as described above with respect to FIG. 25A. Vehicle 200 can then determine an expected distance 2540 based on a lateral distance between the target trajectory 2555 and the mapped lane marks 2550 corresponding to the longitudinal position 2520. The lateral position identification of vehicle 200 can be corrected or adjusted by comparing the actual distance 2530 measured using the captured image with the expected distance 2540 from the model.

[0340] FIG. 26A is a flowchart showing an exemplary process 2600A for mapping lane marks for use in autonomous vehicle navigation according to the disclosed embodiments. At step 2610, process 2600A may include receiving two or more location identifiers associated with the detected lane marks. For example, step 2610 may be performed by server 1230 or one or more processors associated with the server. The location identifier may include the location in the real-world coordinates of the points associated with the detected lane marks, as described above with respect to FIG. 24E. In some embodiments, the location identifier may also include other data such as additional information regarding the road segment or lane mark. Additional data such as accelerometer data, speed data, landmark data, road geometry or profile data, vehicle position data, ego-motion data, or various other forms of data described above may also be received during step 2610. The location identifier may be generated by vehicles such as vehicles 1205, 1210, 1215, 1220, and 1225 based on images captured by the vehicle. For example, the identifier may be determined based on the capture of at least one image from a camera associated with the host vehicle representing the environment of the host vehicle, the analysis of at least one image to detect lane marks in the environment of the host vehicle, and the analysis of at least one image to determine the position of the detected lane marks relative to the position associated with the host vehicle. As described above, the lane marks may include various different mark types, and the location identifier may correspond to various points associated with the lane marks. For example, if the detected lane mark is part of a dashed line marking a lane boundary, the points may correspond to the detected corners of the lane mark. If the detected lane mark is part of a solid line marking a lane boundary, the points may correspond to the detected ends of the lane mark at various intervals, as described above. In some embodiments, the points may correspond to the centerline of the detected lane mark, as shown in FIG. 24C, or to at least one of the vertex between two intersecting lane marks and two other points associated with the intersecting lane marks, as shown in FIG. 24D.

[0341] In step 2612, process 2600A may include associating the detected lane marks with the corresponding road segments. For example, server 1230 may analyze the real-world coordinates or other information received during step 2610 and compare the coordinates or other information with the location information stored in the autonomous vehicle road navigation model. Server 1230 may determine the road segment in the model that corresponds to the real-world road segment where the lane marks were detected.

[0342] In step 2614, process 2600A may include updating the autonomous vehicle road navigation model associated with the corresponding road segment based on two or more location identifiers associated with the detected lane marks. For example, the autonomous vehicle road navigation model may be a sparse map 800, and server 1230 may update the sparse map to include or adjust the lane marks mapped to the model. Server 1230 may update the model based on various methods or processes described above with respect to FIG. 24E. In some embodiments, updating the autonomous vehicle road navigation model may include storing one or more indicators of the location of the detected lane marks at their real-world coordinates. The autonomous vehicle road navigation model may include at least one target trajectory that the vehicle follows along the corresponding road segment, as shown in FIG. 24E.

[0343] In step 2616, process 2600A may include distributing the updated autonomous vehicle road navigation model to a plurality of autonomous vehicles. For example, server 1230 may distribute the updated autonomous vehicle road navigation model to vehicles 1205, 1210, 1215, 1220, and 1225 that may use the model for navigation. The autonomous vehicle road navigation model may be distributed via wireless communication path 1235 through one or more networks (e.g., via a cellular network and / or the Internet, etc.), as shown in FIG. 12.

[0344] In some embodiments, lane marks can be mapped using data received from multiple vehicles via cloud sourcing techniques or the like, as described above with respect to FIG. 24E. For example, process 2600A can include receiving a first communication from a first host vehicle that includes a location identifier associated with a detected lane mark, and receiving a second communication from a second host vehicle that includes an additional location identifier associated with the detected lane mark. For example, the second communication can be received from a subsequent vehicle traveling on the same road segment, i.e., the same vehicle traveling subsequently along the same road segment. Process 2600A can further include refining the determination of at least one location associated with the detected lane mark based on the location identifier received in the first communication and the additional location identifier received in the second communication. This can include using the average of multiple location identifiers and / or excluding "ghost" identifiers that may not reflect the real-world location of the lane mark.

[0345] FIG. 26B is a flowchart showing an exemplary process 2600B for autonomously navigating a host vehicle along a road segment using the mapped lane marks. Process 2600B may be executed, for example, by the processing unit 110 of the autonomous vehicle 200. In step 2620, process 2600B may include receiving an autonomous vehicle road navigation model from a server-based system. In some embodiments, the autonomous vehicle road navigation model may include a target trajectory of the host vehicle along the road segment and position identifiers associated with one or more lane marks associated with the road segment. For example, vehicle 200 may receive a sparse map 800 or another road navigation model developed using process 2600A. In some embodiments, the target trajectory may be represented as a three-dimensional spline, for example, as shown in FIG. 9B. As described above with respect to FIGS. 24A-24F, the position identifiers may include the position in the real-world coordinates of points associated with the lane marks (e.g., corner points of a dashed lane mark, end points of a solid lane mark, vertices between two intersecting lane marks and other points associated with the intersecting lane marks, centerlines associated with the lane marks, etc.).

[0346] In step 2621, process 2600B may include receiving at least one image representing the environment of the vehicle. The image may be received from an image capture device of the vehicle via image capture devices 122 and 124 included in the image acquisition unit 120, etc. The image may include an image of one or more lane marks, similar to the image 2500 described above.

[0347] In step 2622, process 2600B may include determining the longitudinal position of the host vehicle along the target trajectory. As described above with respect to FIG. 25A, this may be based on other information (e.g., landmarks, etc.) within the captured image or by the dead reckoning of the vehicle between the detected landmarks.

[0348] In step 2623, process 2600B may include determining an estimated lateral distance to a lane mark based on the determined longitudinal position of the host vehicle along the target trajectory and based on two or more position identifiers associated with at least one lane mark. For example, vehicle 200 may use sparse map 800 to determine the estimated lateral distance to a lane mark. As shown in FIG. 25B, the longitudinal position 2520 along the target trajectory 2555 may be determined in step 2622. Using sparse map 800, vehicle 200 may determine an estimated distance 2540 to the mapped lane mark 2550 corresponding to the longitudinal position 2520.

[0349] In step 2624, process 2600B may include analyzing at least one image to identify at least one lane mark. Vehicle 200 may use various image recognition techniques or algorithms to identify lane marks within the image, as described above. For example, lane mark 2510 may be detected via image analysis of image 2500, as shown in FIG. 25A.

[0350] In step 2625, process 2600B may include determining an actual lateral distance to at least one lane mark based on the analysis of at least one image. For example, the vehicle may determine a distance 2530 representing the actual distance between the vehicle and lane mark 2510, as shown in FIG. 25A. The angle of the camera, the speed of the vehicle, the width of the vehicle, the position of the camera relative to the vehicle, or various other factors may be considered in determining distance 2530.

[0351] In step 2626, process 2600B may include determining an autonomous steering operation of the host vehicle based on the difference between the predicted lateral distance to at least one lane mark and the determined actual lateral distance to at least one lane mark. For example, as described above with respect to FIG. 25B, vehicle 200 may compare actual distance 2530 with predicted distance 2540. The difference between the actual distance and the predicted distance may indicate the error (and its magnitude) between the actual position of the vehicle and the target trajectory that the vehicle is following. Accordingly, the vehicle may determine an autonomous steering operation or other autonomous operation based on that difference. For example, as shown in FIG. 25B, if actual distance 2530 is shorter than predicted distance 2540, the vehicle may determine an autonomous steering operation to turn the vehicle left away from lane mark 2510. Accordingly, the position of the vehicle relative to the target trajectory may be corrected. Process 2600B may be used, for example, to improve the navigation of the vehicle between landmarks.

[0352] Navigation Based on Image Analysis

[0353] As described above, an autonomous vehicle navigation system or a partially autonomous vehicle navigation system may collect information regarding current conditions, infrastructure, objects, etc. in the vehicle's environment depending on sensor inputs. Based on the collected information, the navigation system may determine one or more navigation actions to take (e.g., based on the application of one or more driving policies to the collected information) and may perform one or more navigation actions via the actuation systems available in the vehicle.

[0354] In some cases, sensors for collecting information about the vehicle's environment may include one or more cameras such as image capture devices 122, 124, and 126 as described above. Each frame captured by the relevant camera can be analyzed by one or more components of the vehicle navigation system to provide the desired functionality of the vehicle navigation system. For example, the captured images can be provided to a separate module responsible for detecting specific objects or features in the environment and navigating in the presence of the detected objects or features. For example, the navigation system may include a separate module for detecting one or more of pedestrians, other vehicles, objects in the road, road boundaries, road markings, traffic lights, traffic signs, parked vehicles, lateral movement of nearby vehicles, opening of doors of parked vehicles, wheel / road boundaries of nearby vehicles, rotation associated with detected wheels of nearby vehicles, road surface conditions (e.g., wet, snow-covered, frozen, gravel-covered, etc.), and the like.

[0355] Each module or function may include the analysis of a provided image (e.g., each captured frame) to detect the presence of specific features on which the logic of that module or function is based. Further, various image analysis techniques may be used to optimize the computational resources for processing and analyzing the captured images. For example, in the case of a module responsible for detecting the opening of a door, the image analysis technique associated with that module may include scanning the image pixels to determine whether there is one associated with a parked car. If no pixels representing a parked car are identified, the work of the module may end in relation to a particular image frame. However, if a parked car is identified, the module may focus on identifying which of the pixels associated with the parked car represent the end of the door (e.g., the end of the rearmost door). These pixels can be analyzed to determine whether there is evidence that the door is open or fully closed. These pixels and the characteristics of the associated door end can be compared across multiple image frames to further confirm whether the detected door is in an open state. In any case, a particular function or module may need to focus only on a portion of the captured image frames to provide the desired function.

[0356] Nevertheless, various image analysis techniques can be used to streamline the analysis in order to efficiently utilize the available computing resources. However, parallel processing of image frames across multiple functions / modules can involve significant use of computing resources. In fact, if each module / function is responsible for its own image analysis to identify the features it depends on, all additional modules / functions added to the navigation system can add image analysis components (e.g., components associated with examining and classifying all pixels). Assuming that each captured frame contains millions of pixels, multiple frames are captured per second, and the frames are supplied to dozens or hundreds of different modules / functions for analysis, the computing resources used to perform the analysis at a speed suitable for driving (e.g., at least as fast as the frames are captured) across many functions / modules can be substantial. In many cases, due to the computational requirements of the image analysis phase, taking into account hardware constraints and the like, the number of functions or features that the navigation system can provide may be limited.

[0357] The described embodiments include an image analysis architecture for addressing these issues in a vehicle navigation system. For example, the described embodiments can include a unified image analysis framework that can remove the burden of image analysis from individual navigation system modules / functions. Such an integrated image analysis framework can include, for example, a single image analysis layer that receives the captured image frames as input, analyzes and characterizes the pixels associated with the captured image frames, and provides the characterized image frames as output. The characterized image frames can then be supplied to multiple different functions / modules of the navigation system to generate and implement appropriate navigation actions based on the characterized image frames.

[0358] In some embodiments, each pixel of an image can be analyzed to determine whether the pixel is associated with a particular type of object or feature within the environment of the host vehicle. For example, each pixel or a portion of the image within the image can be analyzed to determine whether they are associated with another vehicle within the environment of the host vehicle. Additional information can be determined for each pixel, such as whether the pixel corresponds to an end of a vehicle, the surface of a vehicle, etc. Such information can enable more accurate identification and proper orientation of the bounding box that can be shown around the vehicle detected for navigation purposes.

[0359] FIG. 27 shows an example of an analysis performed on pixels associated with a target vehicle according to the disclosed embodiments. FIG. 27 can represent an image or a portion of an image captured by an image capture device such as the image acquisition unit 120. The image can include a representation of a vehicle 2710 within the environment of a host vehicle such as the vehicle 200 described above. The image can be composed of a plurality of individual pixels such as pixels 2722 and 2724. The navigation system (e.g., the processing unit 110) can analyze each pixel to determine whether the pixel is associated with the target vehicle. As used herein, the target vehicle can refer to a vehicle within the environment of the host vehicle that may be navigated with respect to the host vehicle. This can exclude, for example, a vehicle being transported on a trailer or carrier, a reflection of a vehicle, or other representations of a vehicle that can be detected in the image, as discussed in more detail below. The analysis can be performed on all pixels of the captured image or on all pixels within a candidate region identified for the captured image.

[0360] The navigation system can be further configured to analyze each pixel to determine other relevant information. For example, each pixel determined to be associated with the target vehicle can be analyzed to determine which part of the vehicle the pixel is associated with. This can include determining whether the pixel is associated with an end of the vehicle. For example, as shown in FIG. 27, vehicle 2710 can be associated with an end 2712 represented in the image. When analyzing pixel 2724, the navigation system can determine that pixel 2724 is a boundary pixel and thus includes end 2712. In some embodiments, the pixels can also be analyzed to determine whether they are associated with the surface of the vehicle. The terms boundary, end, and surface are defined by the navigation system and relate to a virtual shape that at least partially adjoins or surrounds an object of interest, in this example vehicle 2710. The shape can take any form and can be fixed or specific to the type or class of the object. Typically, by way of example, the shape can include a rectangle that fits snugly around the contour of the object of interest. In other examples, for at least some object classes, a 3D box and each surface of the box can tightly bound the corresponding surface of a 3D object within the environment of the host vehicle. In this particular example, the 3D box bounds vehicle 2710.

[0361] Accordingly, for example, the navigation system may analyze pixel 2722 and determine that pixel 2722 is located on the surface 2730 of the vehicle 2710. The navigation system may further determine one or more estimated distance values from pixel 2722. In some embodiments, one or more distances from pixel 2722 to the end of surface 2730 may be estimated. In the case of a 3D box, the distance may be determined from a given pixel to the end of the surface of the 3D box that forms the boundary of that pixel. For example, the navigation system may estimate the distance 2732 to the side surface of surface 2730 and the distance 2734 to the lower end of surface 2730. The estimation may be based on the position determined to be represented by the portion 2722 of the vehicle and the typical dimensions from that portion of the vehicle to the end of surface 2730. For example, pixel 2722 may correspond to a particular portion of the bumper recognized by the system. Other pixels, such as pixels representing license plates, taillights, tires, exhaust pipes, etc., may be associated with different estimated distances. Although not shown in FIG. 27, various other distances may be estimated, including the distance to the upper end of surface 2730, the distance at the front or rear end (depending on the orientation of vehicle 2710), the distance to the end 2712 of vehicle 2710, etc. The distance may be measured in any unit suitable for analysis by the navigation system. In some embodiments, the distance may be measured in pixel units relative to the image. In other embodiments, the distance may represent the actual distance measured relative to the target vehicle (e.g., centimeters, meters, inches, feet, etc.).

[0362] The processing unit 110 may provide the pixel-based analysis described above using any suitable method. In some embodiments, the navigation system may include a trained neural network that characterizes individual pixels within an image according to a neural network training protocol. The training dataset used to train the neural network may include a plurality of captured images including representations of vehicles. Each pixel of the image may be examined and characterized according to a predetermined set of characteristics that the neural network is to recognize. For example, each pixel may be classified to indicate whether it is part of the representation of the target vehicle, whether the end of the target vehicle is included, the surface of the target vehicle, the measured distance from the pixel to the end of the surface (which may be measured in pixels or real-world distances), or whether it represents other relevant information. The designated image can be used as the training dataset for the neural network.

[0363] Using the resulting trained model, the information associated with the pixel or cluster of pixels can be analyzed to identify whether the pixel represents the target vehicle, whether it is on the surface or end of the vehicle, the distance to the end of the surface, or other information. Although the system has been described throughout this disclosure as a neural network, other various machine learning algorithms may be used, including logistic regression, linear regression, regression, random forest, K-nearest neighbor (KNN) model, K-Means model, decision tree, cox proportional hazards regression model, naive Bayes model, support vector machine (SVM) model, gradient boosting algorithm, deep learning model, or any suitable form of machine learning model or algorithm.

[0364] Based on the mapping of each pixel to the boundary or surface of the target vehicle 2710, the navigation system may more accurately determine the boundary of the vehicle 2710, whereby the host vehicle may accurately determine one or more appropriate navigation operations. For example, the system may be able to determine the complete boundary of the vehicle represented by the end 2712. In some embodiments, the navigation system may determine a bounding box 2720 of the vehicle that may represent the boundary of the vehicle in the image. The bounding box 2720 determined based on the analysis of each pixel may more accurately define the boundary of the vehicle 2710 than conventional object detection methods.

[0365] Furthermore, the system may determine a boundary having an orientation that more accurately represents the orientation of the target vehicle. For example, based on the combined analysis of pixels (e.g., pixel 2722) within the surface 2730 of the vehicle 2710, the end of the surface 2730 may be estimated. By identifying the discrete surfaces of the vehicle 2710 represented in the image, the system may more accurately orient the bounding box 2720 to correspond to the orientation of the vehicle 2710. The improved accuracy of the orientation of the bounding box may improve the system's ability to determine appropriate navigation operation responses. For example, in the case of a vehicle detected from a side-facing camera, an inappropriate orientation of the bounding box may inappropriately indicate that the vehicle is traveling towards the host vehicle (e.g., an intrusion scenario). The disclosed techniques may be particularly beneficial in situations where the target vehicle occupies much, if not all, of the captured image (e.g., situations where at least one end of the vehicle is absent to suggest the orientation of the bounding box), situations where reflections of the target vehicle are included in the image, situations where the vehicle is being carried on a trailer or carrier, etc.

[0366] FIG. 28 is a diagram of an exemplary image 2800 that includes a partial representation of a vehicle 2810 according to the disclosed embodiment. As shown, at least one end (or a portion of an end) of the vehicle 2810 may be excluded from the image 2800. In some embodiments, this may be due to a portion of the vehicle 2810 being outside the field of view of the camera. In other embodiments, a portion of the vehicle 2810 may be blocked from view by, for example, a building, a plant, another vehicle, etc. The configuration of the image 2800 is provided as an example, but in some embodiments, the vehicle 2810 may occupy a larger portion of the image 2800, such that one or more entire ends are excluded.

[0367] When using conventional object detection techniques, the detected boundaries for vehicle 2810 can be inaccurate because the ends not included in the image may be needed to determine the shape and / or orientation of the vehicle. This can be particularly true for images where the vehicle occupies most or all of the image. Using the techniques disclosed herein, each pixel associated with vehicle 2810 within image 2800 can be analyzed to determine bounding box 2820 regardless of whether vehicle 2810 is fully represented within the image. For example, a trained neural network model can be used to determine that pixel 2824 includes an end of vehicle 2810. The trained model can also determine that pixel 2822 is on the surface of vehicle 2810, similar to pixel 2722. The system can also determine one or more distances from pixel 2822 to the ends of the surface of the vehicle, similar to distances 2732 and 2734 described above. In some embodiments, this estimated distance information can be used to define the boundaries of the vehicle that are not included in the image. For example, pixel 2822 can be analyzed to determine an estimated distance from pixel 2822 to the end of vehicle 2810 that extends beyond the end of the image frame. Thus, an accurate boundary of vehicle 2810 can be determined based on an analysis of the combination of pixels associated with vehicle 2810 that appears within the image. This information can be used to determine the navigation operation of the host vehicle. For example, the unseen rear end of a vehicle (e.g., a truck, bus, trailer, etc.) can be determined based on a portion of the vehicle within the image frame. Thus, the navigation system can estimate the clearance required for the target vehicle to determine whether the host vehicle needs to brake or decelerate, move into an adjacent lane, increase speed, etc.

[0368] In some embodiments, the disclosed techniques can be used to determine a bounding box representing a vehicle that is not the target vehicle and, thus, should not be associated with the bounding box for purposes of navigation determination. In some embodiments, this can include a vehicle that is being towed or carried by another vehicle. FIG. 29 is a diagram of an exemplary image 2900 showing a vehicle on a carrier, according to the disclosed embodiments. Image 2900 can be captured by a camera of a host vehicle, such as camera 120 of vehicle 200, as described above. In the example shown in FIG. 29, image 2900 can be a side view image taken by a camera disposed on the side of vehicle 200. Image 2900 can include a carrier vehicle 2910 that can carry one or more vehicles 2920 and 2930. Carrier vehicle 2910 is shown as an automobile transport trailer, but various other vehicle carriers can be identified. For example, carrier vehicle 2910 can include a flatbed trailer, a tow truck, a single car trailer, an incline car carrier, a gooseneck trailer, a drop deck trailer, a wedge trailer, an automobile transport train vehicle, or any other vehicle for transporting another vehicle.

[0369] Using the techniques described above, each pixel in the image can be analyzed to identify the boundaries of the vehicle. For example, the pixels associated with carrier vehicle 2910 can be analyzed to determine the boundaries of carrier vehicle 2910 as described above. The system can also analyze the pixels associated with vehicles 2920 and 2930. Based on the analysis of the pixels, the system can determine that vehicles 2920 and 2930 are not the target vehicles and thus should not be associated with the bounding box. This analysis can be performed in various ways. For example, the trained neural network described above can be trained using a set of training data that includes images of the vehicles being transported. The pixels associated with the vehicles being transported in the image can be designated as either the vehicles being transported or non-target vehicles. Thus, the trained neural network model can determine, based on the pixels of image 2900, that vehicles 2920 and 2930 are being transported by carrier vehicle 2910. For example, the neural network can be trained such that the pixels within vehicle 2920 or 2930 are associated with the ends of carrier vehicle 2910 rather than the ends of vehicle 2920 or 2930. Thus, the boundaries of the vehicles being transported cannot be determined. Vehicles 2920 and 2930 can also be identified as the vehicles being transported using other techniques, for example, based on the position of the vehicles relative to vehicle 2910, the orientation of the vehicles, the position of the vehicles relative to other elements in the image, etc.

[0370] In some embodiments, the system may also determine that a reflection of a vehicle in the image should not be considered the target vehicle, and thus may not determine the boundaries of the reflection. FIGS. 30A and 30B show exemplary images 3000A and 3000B that include reflections of vehicles according to the disclosed embodiments. In some embodiments, the reflection may be based on the surface of the road. For example, as shown in FIG. 30A, image 3000A may include a vehicle 3010 traveling on or at least partially on a reflective surface such as a wet road. Accordingly, image 3000A may also include a reflection 3030 of vehicle 3010. Such reflections can be a problem in an autonomous vehicle system because, in some cases, reflection 3030 can be interpreted as the target vehicle by the system. Accordingly, reflection 3030 is associated with its own bounding box and can be considered in vehicle navigation decisions. For example, host vehicle 200 may determine that reflection 3030 is much closer to vehicle 200 than vehicle 3010, thereby causing vehicle 200 to perform unnecessary navigation operations (e.g., applying brakes, executing a lane change, etc.).

[0371] Using the disclosed method, the pixels associated with reflection 3030 can be analyzed to determine that reflection 3030 is not the target vehicle, and thus the boundary of the reflected vehicle should not be determined. This can be performed in a similar manner as the method described above for vehicles on the carrier. For example, a neural network model can be trained using an image that includes the reflection of the vehicle such that the individual pixels associated with the reflection can be identified as either indicative of the reflection or not indicative of the target vehicle. Thus, the trained model can be capable of distinguishing the pixels associated with the target vehicle such as vehicle 3010 and reflection 3030. In some embodiments, reflection 3030 can be identified as a reflection based on its orientation, its position relative to vehicle 3010, its position relative to other elements of image 3000A, etc. The system can determine the bounding box 3020 associated with vehicle 3010, but cannot determine a bounding box or other boundary associated with reflection 3030. Reflection 3030 was described above as appearing on a wet road surface, but reflections can also appear on other surfaces such as a metal or other reflective surface, a mirage reflection due to a heated road surface, etc.

[0372] In addition to reflections on the road, reflections of the vehicle can be detected based on reflections on other surfaces. For example, as shown in FIG. 30B, a reflection of the vehicle can be displayed within an image on the surface of another vehicle. Image 3000B can represent an image captured by a camera of vehicle 200, such as a side view camera. Image 3000B can include, in this example, a representation of another vehicle 3050 that can be traveling alongside vehicle 200. The surface of vehicle 3050 can be at least partially reflective such that a reflection 3060 of a second vehicle can appear in image 3000B. Reflection 3060 can be a reflection of host vehicle 200 that appears on the surface of vehicle 3050, or a reflection of another target vehicle. Similar to reflection 3030, the system can analyze the pixels associated with reflection 3060 to determine that reflection 3060 does not represent a target vehicle, and thus, the boundaries of reflection 3060 should not be determined. For example, a neural network can be trained such that the pixels within reflection 3060 are associated with the edges of vehicle 3050 rather than the edges of the reflected vehicle representation. Thus, the boundaries of reflection 3060 cannot be determined. Reflection 3060 can be ignored for the purpose of determining navigation actions once it is identified as a reflection. For example, if the movement of the reflection appears to be moving towards the host vehicle, the host vehicle may not apply the brakes or perform other navigation actions that it might otherwise perform if that movement was being executed by a target vehicle. Reflection 3060 is shown in FIG. 30B as appearing on the side of a tanker truck, but can also appear on other surfaces such as a shiny painted surface of another vehicle (e.g., a door, side panel, bumper, etc.), a chrome surface, a glass surface of another vehicle (e.g., a window, etc.), a building (e.g., building windows, metal surfaces, etc.), or other reflective surfaces that can reflect an image of the target vehicle.

[0373] FIG. 31A is a flowchart showing an exemplary process 3100 for navigating a host vehicle based on analysis of pixels in an image according to the disclosed embodiment. Process 3100 can be executed by at least one processing device, such as processing unit 110, as described above. Throughout this disclosure, it should be understood that the term "processor" is used as an abbreviation for "at least one processor". In other words, a processor can include one or more structures that perform logical operations, regardless of whether such structures are arranged, connected, or distributed. In some embodiments, a non-transitory computer-readable medium can include instructions that, when executed by a processor, cause the processor to execute process 3100. Further, process 3100 is not necessarily limited to the steps shown in FIG. 31A, and any steps or processes of various embodiments described throughout this disclosure can also be included in process 3100, including those described above with respect to FIGS. 27-30B.

[0374] In step 3110, process 3100 can include receiving at least one captured image representing the environment of the host vehicle from a camera of the host vehicle. For example, image acquisition unit 120 can capture one or more images representing the environment of host vehicle 200. The captured image can correspond to an image as described above with respect to FIGS. 27-30B.

[0375] In step 3120, process 3100 may include analyzing one or more pixels of at least one captured image to determine whether one or more pixels represent at least a portion of a target vehicle. For example, pixels 2722 and 2724 may be analyzed as described above to determine that the pixels represent a portion of vehicle 2710. In some embodiments, the analysis may be performed on all pixels of the captured image. In other embodiments, the analysis may be performed on a subset of the pixels. For example, the analysis may be performed on all pixels of a target vehicle candidate region identified in relation to the captured image. Such a region may be determined, for example, using process 3500 described in detail below. The analysis may be performed using various techniques, including the trained system described above. Thus, a trained system that may include one or more neural networks may perform at least a portion of the analysis of one or more pixels.

[0376] In step 3130, process 3100 may include determining one or more estimated distance values from one or more pixels to at least one end of the surface of the target vehicle in the case of pixels determined to represent at least a portion of the target vehicle. For example, as described above, processing unit 110 may determine that pixel 2722 represents surface 2730 of vehicle 2710. Accordingly, one or more distances, such as distances 2732 and 2734, to the ends of surface 2730 may be determined. For example, the distance value may include a distance from a particular pixel to at least one of a front end, a rear end, a side end, a top end, or a bottom end of the target vehicle. The distance value may be measured in pixel units based on the image or in real-world distances with respect to the target vehicle. In some embodiments, process 3100 may further include determining whether one or more pixels include boundary pixels that include a representation of at least a portion of at least one end of the target vehicle. For example, pixel 2724 may be identified as a pixel that includes a representation of end 2712 as described above.

[0377] In step 3140, process 3100 may include generating at least a portion of a boundary for a target vehicle based on an analysis of one or more pixels, including one or more determined distance values associated with the one or more pixels. For example, step 3140 may include determining a portion of the boundary represented by end 3712, as shown in FIG. 27. This may be determined based on a combined analysis of all pixels associated with target vehicle 2710 within the image. For example, end 3712 may be estimated based on a combination of estimated distance values generated based on pixels included in the surface of the vehicle, along with the pixels identified as including end 3712. In some embodiments, a portion of the boundary may include at least a portion of a bounding box, such as bounding box 2720.

[0378] Process 3100 may be executed in certain scenarios to improve the accuracy of the determined boundary of the target vehicle. For example, in some embodiments, as shown in FIG. 28, at least a portion of at least one end (or one or more ends) of the target vehicle may not be represented in the captured image. Using process 3100, the boundary of the target vehicle (e.g., vehicle 2810) may be determined based on the pixels associated with the target vehicle that appear within the image. In some embodiments, process 3100 may further include determining, based on an analysis of the captured image, whether the target vehicle is being carried by another vehicle or trailer, as described above with respect to FIG. 29. Accordingly, processing unit 110 may not be able to determine the boundary of the carried vehicle. Similarly, process 3100 may include determining, based on an analysis of the captured image, whether the target vehicle is included in the representation of a reflection of at least one image, as described above with respect to FIGS. 30A and 30B. In such embodiments, processing unit 110 may not be able to determine the boundary of the vehicle reflection.

[0379] In some embodiments, process 3100 may include additional steps based on the above analysis. For example, process 3100 may include determining the orientation of at least a portion of the generated boundary for the target vehicle. As described above, the orientation may be determined based on a combined analysis of each pixel associated with the target vehicle, including one or more identified surfaces associated with the target vehicle. The determined orientation may indicate an action being performed or planned to be performed (or future action or state) by the target vehicle. For example, the determined orientation may indicate a lateral movement (e.g., a lane change action) of the target vehicle with respect to the host vehicle or an action by the target vehicle toward the path of the host vehicle. Process 3100 may further include determining a navigation action of the host vehicle based on the determined orientation of at least a portion of the boundary and causing the vehicle to perform the determined navigation action. For example, the determined navigation action may include a merging action, a braking action, an accelerating action, a lane change action, a sharp turn, or other avoidance actions, etc. In some embodiments, process 3100 may further include determining the distance to the host vehicle of a portion of the boundary. This may include determining the position of a portion of the boundary within the image and estimating the distance to the host vehicle based on the position. Various other techniques described throughout the present disclosure may also be used to determine the distance. Process 3100 may further include causing the vehicle to perform a navigation action based at least on the determined distance.

[0380] In some embodiments, process 3100 may include determining information regarding a target vehicle based on the analyzed pixels. For example, process 3100 may include determining the type of the target vehicle and outputting the type of the target vehicle. Such types of target vehicles may include at least one of a bus, a truck, a bicycle, a motorcycle, a van, a car, a construction vehicle, an emergency vehicle, or other types of vehicles. The type of the target vehicle may be determined based on the size of a part of the boundary. For example, the number of pixels included within the boundary may indicate the type of the target vehicle. In some embodiments, the size may be ...

Claims

1. A navigation system for a host vehicle, the navigation system comprising: at least one processor having circuitry, and a memory, the memory, when executed by the circuitry, causing the at least one processor to: receive at least one image captured by a camera of the host vehicle, wherein the at least one image represents the environment of the host vehicle; analyze one or more pixels of the at least one image to determine whether the one or more pixels represent at least a part of a target vehicle, and for pixels determined to represent at least a part of the target vehicle, determine one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle; generate at least a part of a boundary for the target vehicle based on the analysis of the one or more pixels, including the one or more determined estimated distance values associated with the one or more pixels, wherein the boundary represents an outer boundary of the representation of the target vehicle in the at least one image; A navigation system comprising instructions to perform the above.

2. The navigation system according to claim 1, wherein the one or more estimated distance values are measured in pixel units.

3. The navigation system according to claim 1 or 2, wherein the one or more estimated distance values correspond to real-world distances measured with respect to the target vehicle.

4. The navigation system according to any one of claims 1 to 3, wherein the one or more estimated distance values include the distance from a specific pixel to at least one of a front end, a rear end, a side end, an upper end, or a lower end of the target vehicle.

5. The navigation system according to any one of claims 1 to 4, wherein the at least one processor is further programmed to determine the orientation of the at least a part of the boundary generated for the target vehicle.

6. The navigation system according to claim 5, wherein the at least one processor is further programmed to determine a navigation operation of the host vehicle based on the determined orientation of the at least a part of the boundary and cause the host vehicle to perform the determined navigation operation.

7. The navigation system according to claim 5 or 6, wherein the determined direction indicates an operation of the target vehicle toward the route of the host vehicle.

8. The navigation system according to claim 5 or 6, wherein the determined direction indicates a lateral movement of the target vehicle with respect to the host vehicle.

9. The navigation system according to any one of claims 1 to 8, wherein the at least one processor is further programmed to determine whether the one or more pixels include boundary pixels that include a representation of at least a part of at least one end of the target vehicle.

10. The navigation system according to any one of claims 1 to 9, wherein the analysis is performed on all pixels of the captured image.

11. The navigation system according to any one of claims 1 to 9, wherein the analysis is performed on all pixels of a target vehicle candidate region identified in the captured image.

12. The navigation system according to any one of claims 1 to 11, wherein the part of the boundary includes at least a part of a bounding box.

13. The navigation system according to any one of claims 1 to 12, wherein the at least one processor determines a distance from the part of the boundary to the host vehicle and is further programmed to cause the host vehicle to perform a navigation operation based at least on the determined distance.

14. The navigation system according to any one of claims 1 to 13, wherein a trained system performs at least a part of the analysis of the one or more pixels.

15. The navigation system according to claim 14, wherein the trained system includes one or more neural networks.

16. The navigation system according to any one of claims 1 to 15, wherein at least a part of at least one end of the target vehicle is not represented in the captured image.

17. The one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle include at least one estimated distance value from the one or more pixels to one or more ends of the target vehicle not represented in the captured image. The navigation system according to any one of claims 1 to 6.

18. The at least one processor is further programmed to determine, based on an analysis of the captured image, whether the target vehicle is being carried by another vehicle or a trailer. The navigation system according to any one of claims 1 to 17.

19. The at least one processor is further programmed not to determine the boundaries of the vehicle being carried. The navigation system according to claim 18.

20. The at least one processor is further programmed to determine, based on an analysis of the captured image, whether the target vehicle is included in the representation of the reflection of the at least one image. The navigation system according to any one of claims 1 to 19.

21. The at least one processor is further programmed not to determine the boundaries of the vehicle reflection. The navigation system according to claim 20.

22. The at least one processor is further programmed to output the type of the target vehicle. The navigation system according to any one of claims 1 to 21.

23. The type of the target vehicle is based at least on the size of at least a part of the boundary. The navigation system according to claim 22.

24. The type of the target vehicle is at least partially based on the number of pixels included within the boundary. The navigation system according to claim 22.

25. The type of the target vehicle includes at least one of a bus, a truck, a bicycle, a motorcycle, or a passenger car. The navigation system according to any one of claims 22 to 24.

26. A navigation system for a host vehicle, the navigation system comprising at least one processor having circuitry and a memory, the memory, when executed by the circuitry, causing the at least one processor to Receiving a first image captured by a camera of the host vehicle, where the first image represents the environment of the host vehicle; Analyzing one or more pixels of the first image to determine whether the one or more pixels represent at least a part of a target vehicle, and for the pixels determined to represent at least a part of the target vehicle, determining one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle; Generating at least a part of a first boundary for the target vehicle based on the analysis of the one or more pixels of the first image, including the one or more determined estimated distance values associated with the one or more pixels of the first image, where the first boundary represents the outer boundary of the representation of the target vehicle in the first image; Receiving a second image captured by a camera of the host vehicle, where the second image represents the environment of the host vehicle; Analyzing one or more pixels of the second image to determine whether the one or more pixels represent at least a part of the target vehicle, and for the pixels determined to represent at least a part of the target vehicle, determining one or more estimated distance values from the one or more pixels to at least one end of the surface of the target vehicle; Generating at least a part of a second boundary for the target vehicle based on the analysis of the one or more pixels of the second image, including the one or more determined estimated distance values associated with the one or more pixels of the second image, and based on the first boundary, where the second boundary represents the outer boundary of the representation of the target vehicle in the second image; A navigation system including instructions to cause the above to be performed.

27. The navigation system according to claim 26, wherein a trained system executes at least a part of the analysis of the one or more pixels.

28. The navigation system according to claim 27, wherein the trained system includes one or more neural networks.

29. The navigation system according to any one of claims 26 to 28, wherein the first boundary includes at least a part of a first bounding box, and the second boundary includes at least a part of a second bounding box.

Images