METHOD AND DEVICE FOR TRAFFIC LIGHT POSITIONING AND MAPPING USING CROWD-SENSED DATA

Abstract

A method for locating a traffic light (202) on a carrier vehicle (10), comprising: Acquiring image data at an intersection that includes the traffic light (202) on a probe vehicle (204) when the probe vehicle (204) is at the intersection; Identifying the traffic light (202) at the intersection from the image data; Create, on a remote processor (206), an observed node in a digital map (220) corresponding to the traffic light (202); Updating, on the remote processor (206), a mapped position of a mapped node within the digital map (220) based on an observed position of the observed node, wherein the mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map (220); and Locating the traffic light (202) at the carrier vehicle (10) using the mapped node of the digital map (220) when the carrier vehicle (10) is at the intersection.

Description

INTRODUCTION

[0001] The disclosure of the subject matter relates to the determination of the location or position of a traffic light using sensors of a vehicle and, in particular, to the use of data from the set of data relating to the location of the traffic light in order to reduce the computational effort on the vehicle for locating the traffic light.

[0002] When an autonomous or semi-autonomous vehicle approaches an intersection, it is expected to comply with the intersection's traffic rules, such as stopping at a red light, braking at a yellow light, and so on. In a traffic light recognition system, the vehicle first receives an image of the intersection and locates the traffic light within the image using various computational methods. These methods can be very time-consuming and computationally intensive. Simultaneously, it is necessary to identify a traffic light and its state within the time it takes the vehicle to reach the intersection. Therefore, it is desirable to provide a system that reduces the time and computational effort required by the autonomous vehicle to identify the location and current state of a traffic light.

[0003] DE 10 2018 007 962 A1 describes a method for detecting traffic light positions, wherein a traffic scene at a traffic light intersection is captured by a multiple fleet vehicles using a camera, traffic light candidates are identified in the captured traffic scene and uploaded to a central backend, wherein the geometric positions of the traffic light candidates relative to each other are determined in the respective fleet vehicle and uploaded to the backend, wherein, if multiple geometric positions for the same traffic scene are uploaded, a statistical analysis of these multiple uploaded geometric positions is performed in the backend, such that frequently occurring traffic light candidates are interpreted as traffic lights, while rarely occurring traffic light candidates are interpreted as false detections.where a pattern with confidence intervals of the traffic light positions and a distance-dependent scaling factor is created from the geometric positions interpreted as traffic lights and assigned to the respective traffic light intersection and a specific direction of travel.

[0004] US 2012 / 0288138A1 describes a method and a system that can determine the location of a vehicle, capture an image using a camera attached to the vehicle, analyze the image in conjunction with the vehicle's location and / or previously collected information about the location of traffic lights or other objects (e.g., traffic signs), and use this analysis to locate a traffic light within the captured image. The signal's position (e.g., a geographic location) can be determined and stored for later use. Identifying the signal can be used to provide an output, such as the signal's status, for example, green light. DESCRIPTION

[0005] According to the invention, a method for locating a traffic light on a carrier vehicle is disclosed. Image data is acquired at an intersection containing the traffic light, and image data is acquired on a probe vehicle when the probe vehicle is at the intersection. The traffic light is identified at the intersection from the image data. On a remote processor, an observed node corresponding to the traffic light is created in a digital map, and the mapped position of this node is updated within the digital map based on the observed position of the node. The mapped position and the observed position of the node are displayed by three-dimensional coordinates within the digital map. The traffic light is located on the carrier vehicle using the mapped node of the digital map when the carrier vehicle is at the intersection.

[0006] In addition to one or more of the features described herein, the method further includes using the mapped node of the digital map to locate the traffic light within the image data of the intersection acquired from the carrier vehicle. The method further includes updating the digital map based on a confidence coefficient associated with the observed node. The method further includes assigning a utility value to the mapped node in the digital map, with the utility value increasing with a number of observations within a selected time period. The method further includes removing the mapped node from the digital map if the utility value of the mapped node falls below a distance threshold.The method further includes determining the location of the carrier vehicle from an observation of the traffic light on the carrier vehicle and the mapped position of the mapped node within the digital map. The method further includes adding a new mapped node to the digital map to represent the traffic light if the observed node associated with the traffic light does not align with the mapped node. The mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map. The probe vehicle transmits a local position of the traffic light to the remote processor, and the remote processor determines the observed position of the observed node from the local position.The probe vehicle also transmits at least one traffic light status, one traffic light signal phase and time, the GPS coordinates of the probe vehicle, the vehicle sensor data and a detection reliability to the remote processor.

[0007] According to the invention, a system for locating a traffic light on a carrier vehicle is disclosed. The system comprises at least one probe vehicle and a remote processor. The at least one probe vehicle is configured to receive image data of an intersection that includes the traffic light when the probe vehicle is located at the intersection, and to identify the traffic light at the intersection from the image data. The remote processor is configured to generate an observed node in a digital map corresponding to the traffic light and to update a mapped position of a mapped node within the digital map based on an observed position of the observed node, wherein the mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map.The carrier vehicle uses the mapped node of the digital map to locate the traffic light when the carrier vehicle is at the intersection.

[0008] In addition to one or more of the features described here, the carrier vehicle uses the mapped node of the digital map to locate the traffic light within the intersection image data acquired by the carrier vehicle. The remote processor is further configured to update the digital map based on a confidence coefficient assigned to the observed node. The remote processor is also configured to assign a utility value to the mapped node in the digital map, with the utility value increasing with a certain number of observations within a selected time period. Finally, the remote processor is configured to remove the mapped node from the digital map when the utility value of the mapped node falls below a distance threshold.The carrier vehicle is configured to determine its location by observing the traffic light on the carrier vehicle and the mapped position of the mapped node within the digital map. The remote processor is further configured to add a new mapped node to the digital map to represent the traffic light if the observed node associated with the traffic light does not align with the mapped node. The mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map. The probe vehicle is further configured to transmit a local position of the traffic light to the remote processor, and the remote processor is further configured to determine the observed position of the observed node from the local position.The probe vehicle is further equipped to transmit at least one traffic light status, one traffic light signal phase and time, the GPS coordinates of the probe vehicle, the vehicle sensor data and a detection reliability to the remote processor.

[0009] The features and advantages mentioned above, as well as other features and advantages of the disclosure, are readily apparent from the following detailed description when considered in conjunction with the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES

[0010] Further features, advantages, and details appear only as examples in the following detailed description, which refers to the figures in which: Fig. Figure 1 shows a semi-autonomous or autonomous vehicle according to an exemplary embodiment; Fig. Figure 2 shows a system for determining the location of a traffic light and for creating a traffic light location map; Fig. Figure 3 shows a flowchart illustrating a procedure performed on a probe vehicle to provide a remote processor with the local position of a traffic light; Fig. 4 shows a top view of the digital map; Fig. Figure 5 shows a flowchart of a two-part matching process for assigning observed nodes to mapped nodes within the digital map of Fig. 4; Fig. Figure 6 illustrates a possible candidate pairing for a matching algorithm based on the observed nodes and the mapped nodes of Fig. 4; Fig. Figure 7 shows a diagram illustrating the development of a utility value for an illustratively mapped node; and Fig. Figure 8 shows an illustrative graph of an average light position based on a selected number of observations or passes through an intersection point. DETAILED DESCRIPTION

[0011] The following description is merely exemplary and is not intended to limit the present disclosure, its application, or its use. It should be understood that throughout the figures, corresponding reference numbers point to identical or equivalent parts and features.

[0012] Fig. Figure 1 shows a vehicle 10 according to an exemplary embodiment. In one exemplary embodiment, the vehicle 10 is a semi-autonomous or autonomous vehicle. In various embodiments, the vehicle 10 includes at least one driver assistance system for both steering and acceleration / deceleration, using information about the driving environment, such as cruise control and lane keeping assist. While the driver can be decoupled from the physical operation of the vehicle 10 by simultaneously removing their hands from the steering wheel and their foot from the pedal, the driver must be ready to assume control of the vehicle.

[0013] In general, a trajectory planning system 100 determines a trajectory plan for the automated driving of the vehicle 10. The vehicle 10 generally consists of a chassis 12, a body 14, front wheels 16, and rear wheels 18. The body 14 is mounted on the chassis 12 and essentially encloses components of the vehicle 10. The body 14 and the chassis 12 can together form a frame. The wheels 16 and 18 are each rotatably coupled to the chassis 12 near their respective corners.

[0014] As shown, the vehicle 10 generally comprises a drive system 20, a transmission system 22, a steering system 24, a braking system 26, a sensor system 28, an actuator system 30, at least one data storage device 32, at least one control unit 34, and a communication system 36. The drive system 20 may, in various embodiments, comprise an internal combustion engine, an electric machine such as a traction motor, and / or a fuel cell drive system. The transmission system 22 is configured to transmit power from the drive system 20 to the vehicle wheels 16 and 18 according to selectable speed ratios. According to various embodiments, the transmission system 22 may comprise a continuously variable automatic transmission, a continuously variable transmission, or another suitable transmission. The braking system 26 is configured to deliver braking torque to the vehicle wheels 16 and 18.The braking system 26 can, in various embodiments, comprise friction brakes, a wire brake, a regenerative braking system such as an electric motor, and / or other suitable braking systems. The steering system 24 influences the position of the vehicle wheels 16 and 18. Although the steering system 24 is shown with a steering wheel for illustrative purposes, in some embodiments considered within the scope of this disclosure it may not include a steering wheel.

[0015] The sensor system 28 comprises one or more sensor devices 40a-40n that detect observable conditions of the external environment and / or the vehicle's internal environment 10. The sensor devices 40a-40n may include, among others, radars, lidar, global positioning systems, optical cameras, digital cameras, thermal imaging cameras, ultrasonic sensors, digital video recorders, and / or other sensors for observing and measuring parameters of the external environment. The sensor devices 40a-40n may further include brake sensors, steering angle sensors, wheel speed sensors, etc., for observing and measuring vehicle internal parameters. The cameras may include two or more digital cameras arranged at a selected distance from one another, the two or more digital cameras being used to obtain stereoscopic images of the environment in order to obtain a three-dimensional image.The actuator system 30 comprises one or more actuator devices 42a-42n that control one or more vehicle features, such as, but not limited to, the drive system 20, the transmission system 22, the steering system 24, and the braking system 26. In various embodiments, the vehicle features may also include internal and / or external vehicle features, such as, but not limited to, doors, a trunk, and cabin features such as air conditioning, music, lighting, etc. (not numbered).

[0016] The at least one controller 34 comprises at least one processor 44 and a computer-readable storage device or medium 46. The at least one processor 44 can be any custom or commercially available processor, a central processing unit (CPU), a graphics processing unit (GPU), an auxiliary processor among several processors connected to the at least one controller 34, a semiconductor-based microprocessor (in the form of a microchip or chipset), a macroprocessor, any combination thereof, or generally any device for executing instructions. The computer-readable storage device or media 46 can, for example, include volatile and non-volatile storage in read-only memory (ROM), random-access memory (RAM), and keep-alive memory (KAM).KAM is a persistent or non-volatile memory that can be used to store various operating variables while at least one processor 44 is switched off. The computer-readable storage device or medium 46 can be implemented using any of the known storage devices such as PROMs (programmable read-only memory), EPROMs (electrical PROMs), EEPROMs (electrically erasable PROMs), flash memory, or any other electrical, magnetic, optical, or combined storage device capable of storing data, some of which represents executable instructions used by the at least one control unit 34 in controlling the vehicle 10.

[0017] The instructions can comprise one or more separate programs, each containing an ordered list of executable instructions for implementing logical functions. When the instructions are executed by the at least one processor 44, they receive and process signals from the sensor system 28, perform logic, calculations, procedures, and / or algorithms for the automatic control of the vehicle 10 components, and generate control signals for the actuator system 30 to automatically control the vehicle 10 components based on the logic, calculations, procedures, and / or algorithms. Although in Fig. Where only one control unit is shown, embodiments of the vehicle 10 may contain any number of control units which communicate via any suitable communication medium or combination of communication media and which cooperate to process the sensor signals, perform logic, calculations, procedures and / or algorithms and generate control signals for automatic control of features of the vehicle 10.

[0018] The communication system 36 is configured to wirelessly transmit information to and from other entities 48, such as, but not limited to, other vehicles (“V2V” communication), infrastructure (“V2I” communication), remote systems, and / or personal devices. In one exemplary embodiment, the communication system 36 is a wireless communication system configured to communicate via a wireless local area network (WLAN) using the IEEE 802.11 standard or using cellular data communication. However, additional or alternative communication methods, such as a dedicated short-range communication (DSRC) channel, are also considered within the scope of this disclosure.DSRC channels refer to one-way or two-way short- to medium-range wireless communication channels specifically designed for use in motor vehicles, as well as a corresponding set of protocols and standards.

[0019] Fig. Figure 2 shows a system 200 for determining the location of a traffic light 202 and for creating a traffic light location map. The system 200 comprises a probe vehicle 204, a remote processor 206, and a communication link 208 between the probe vehicle 204 and the remote processor 206. In various embodiments, the probe vehicle 204 can be an autonomous vehicle, as shown in relation to Fig. 1. The probe vehicle 204 contains at least one image sensor 210, such as a digital camera or a digital video recorder. The at least one image sensor 210 is used to capture or collect one or more images of an intersection as the probe vehicle 204 approaches the intersection. The probe vehicle 204 may also contain various sensors for determining the position, orientation, and / or state of the probe vehicle 204. Such sensors may include a GPS (Global Positioning Satellite) receiver for determining the position of the probe vehicle 204, additional sensors for determining the orientation of the probe vehicle 204, such as pitch, roll, and yaw of the probe vehicle 204, etc. Speed ​​sensors may be used to determine the speed of the probe vehicle 204. These sensors may be connected to a processor of the vehicle via a Controller Area Network (CAN).

[0020] The remote processor 206 can be a cloud processor or any other suitable processor. The remote processor 206 sends and receives data from the probe vehicle 204. The remote processor 206 also maintains two-way communication with other probe vehicles 212.

[0021] As the probe vehicle 204 approaches an intersection, the at least one image sensor 210 captures one or more images of the intersection. The probe vehicle 204 applies various algorithms to the one or more images to identify a traffic light 202 within the one or more images and to determine a local position of the traffic light 202 (i.e., a position of the traffic light 202 within the current frame of reference of the probe vehicle 204). In various embodiments, the local position includes three-dimensional coordinates. Generally, the traffic light 202 occupies a small space within one or more images. Therefore, the method for identifying the traffic light 202 from an image of the at least one image sensor 210 is computationally intensive and time-consuming. Once the probe vehicle 204 has identified the local position of the traffic light 202, the local position, as well as other data (i.e., vehicle speed, GPS coordinates, etc.), are transmitted.) uploaded to the remote processor 206 via communication link 208.

[0022] The remote processor 206 stores a digital map 220, including the data structures belonging to each traffic light. A data structure comprises a node representing a traffic light and the node's position within the digital map. The reference frame of the digital map 220 can be a geocentric standard reference frame. The digital map 220 contains mapped nodes based on data from at least one other probe vehicle and often on data from the mass of the other probe vehicles 212. When the remote processor 206 receives a local position for the traffic light 202 from a probe vehicle 204, the remote processor 206 inserts a node, here referred to as the "observed node," into the digital map 220 at an observed position of the observed node. The remote processor 206 further determines the coordinates of the observed node from the local position provided by the probe vehicle 212.The observed node can then be assigned to or associated with a mapped node based on their proximity. The mapped node has a mapped position within the digital map. The mapped position and the observed position are represented by three-dimensional coordinates within the digital map. The pairing of the observed node with the mapped node is determined by calculating a measure or distance between them (i.e., between the observed position of the observed node and a mapped position of the mapped node). The remote processor 206 executes a position update algorithm to update the mapped position of the mapped node within the digital map 220 using the observed position of the observed node.The remote processor 206 also works with a map maintenance algorithm to update the confidence of the mapped node and to remove the mapped node from the digital map when it is no longer supported by the observations of the probe vehicles.

[0023] As in Fig. As shown in Figure 2, the remote processor 206 can download the digital map 220, or a portion thereof, to a carrier vehicle (i.e., vehicle 10). Vehicle 10 takes one or more pictures of the intersection and the traffic light 202 with its digital camera or a suitable sensor as it approaches the intersection. Using information from the digital map 220, as well as its position and orientation relative to the intersection, vehicle 10 is able to determine or approximate the local position of the traffic light 202 within its one or more pictures in a shorter time than is possible for the probe vehicle 204 (which is done without using the digital map 220). Once the traffic light is within the one or more pictures, the carrier vehicle can determine the state of the traffic light and take appropriate actions to comply with the traffic rules indicated by the state of the traffic light.

[0024] Fig. Figure 3 shows a flowchart 300, which shows a procedure that was carried out on the probe vehicle 204 to provide a local position of a traffic light 202 and additional data to the remote processor 206.

[0025] Box 302 records the sensor data from the probe vehicle 204. This sensor data includes one or more images captured by the at least one image sensor 210, such as the digital camera, as well as various lidar and radar data, and data on the status or position of the probe vehicle. This status data includes GPS data, vehicle speed, vehicle tilt, vehicle rotation, vehicle roll, and vehicle yaw, as well as any other suitable data useful for determining the coordinates of the traffic light.

[0026] In box 304, the sensor data on the probe vehicle 204 are preprocessed. This preprocessing includes processes such as data synchronization and noise removal, e.g., through data interpolation, curve fitting, and other suitable methods.

[0027] In Box 306, traffic light 202 is identified from the sensor data. An object recognition method uses sensor fusion algorithms that combine data from multiple sensors, such as cameras, lidar, radar, etc., to detect traffic light 202. In various embodiments, traffic light 202 is identified or detected across a variety of timeframes of the sensor data.

[0028] In box 308, traffic light 202 is tracked over several time frames of sensor data using one or more object tracking algorithms. During object tracking, the traffic light detected in a first frame is linked to the traffic light detected in a second or subsequent frame.

[0029] Box 310 describes how to determine the local three-dimensional position of the traffic light at probe vehicle 204. This position is derived from the GPS position of probe vehicle 204 and the results of object detection and tracking. The estimation of the local three-dimensional position can utilize one or more Bayesian estimation algorithms, such as particle filters, Kalman filters, etc.

[0030] Box 312 describes the use of various filters to eliminate false positives from traffic lights. False positives include objects such as vehicle taillights, lights on buildings, etc. Illustrative filter conditions include, but are not limited to, the following: a traffic light is more than 5 meters above the ground, or its position is near the center of an intersection.

[0031] In box 314, the remaining local three-dimensional positions are uploaded to the remote processor 206. The uploaded data can include, for example, traffic light positions (latitude, longitude, altitude), traffic light status, traffic light signal phase and time, vehicle GPS track, vehicle sensor data, detection confidence, etc., using a suitable wireless connection such as cellular (4G / 5G), DSRC V2I, WiFi, etc.

[0032] Fig. Figure 4 shows a top view 400 of the digital map 220, which illustrates various mapped nodes, representing human-generated traffic lights, and observed nodes, representing observed traffic lights provided by the probe vehicle 204. Mapped nodes 402, 404, 406, and 408 are shown at various intersections of the digital map 220. Observed nodes 412, 414, and 416 are shown at their respective locations on the digital map 220. As explained below, a most-probable pair is formed between a mapped node and an observed node based on their relative position and / or proximity, and then the mapped position of the mapped node within the digital map 220 is updated based on the observed position of the observed node with which it is paired.

[0033] Fig. Figure 5 shows a flowchart 500 of a two-part matching process for the assignment of observed nodes to mapped nodes within the digital map 220.

[0034] Box 502 identifies all possible candidate pairings. A candidate pairing can be determined by a systematic pairing procedure. A possible candidate pairing has no matches between observed nodes and mapped nodes. This candidate pairing is represented in Eq. (1): No match(V1,V2,V3,M1,M2,M3,M4) where V j an observed node in conjunction with a traffic light, which is observed by probe vehicle 204, and M ia mapped node that is stored in the digital map 220. Another set of candidate pairings includes those in which only one pairing between observed nodes and mapped nodes can be found. Several possible one-match pairings are shown in Eqs. (2) and (3): V1↔M1:(V2,V3,M2,M3,M4) V1↔M2:(V2,V3,M1,M3,M4) For another possible candidate, two matches between observed nodes and mapped nodes can be found, as shown in Eq. (4): V1↔M1,V2↔M2:(V3,M3,M4)

[0035] This process will continue until all candidate pairs have been determined.

[0036] Fig. Figure 6 shows a possible candidate pairing 600, which can be used for the matching algorithm of Box 502 based on the information in Fig. The four observed and mapped nodes shown are shown. As shown in the possible candidate pairing, the first observed node 412 is paired with the second mapped node 404, the second observed node 414 with the fourth mapped node 408, and the third observed node 416 with the first mapped node 402. No observed node appears to be paired with the third mapped node 406.

[0037] Back to Fig. In section 5, box 504, a cost function is calculated for each candidate pairing, as in Eq: f(m1,m2,...,mn)=∑(distance(mi))2nα where m i represents a pairing of a mapped node and an observed node, and distance (m) i ) a distance function between the mapped node and the observed node of the pair (m i ) is. The parameter n is a number of pairings and α is a scaling factor.

[0038] Box 506 returns the candidate pairing with the lowest cost value as the best solution for pairing observed traffic lights with mapped traffic lights.

[0039] Once an optimal or desired solution has been found in field 506, a confidence coefficient β is calculated for each matching pair of solutions. The confidence coefficient β is an estimate of the confidence of the observed node based on various vehicle parameters during the acquisition of the local position by the probe vehicle 204, such as the distance between the probe vehicle and the traffic light, the model of the probe vehicle 204, the quality of the sensor 210, and a history of previous observations by the probe vehicle. For example, the confidence coefficient can be reduced for a probe vehicle 204 that has produced false positives in the past. Regarding the distance between the probe vehicle and the traffic light, the closer the probe vehicle is in general, the more accurate the local position data. However, if the camera is too close to the traffic light (e.g.,(directly below the traffic light), trust in the data may be low.

[0040] The confidence coefficient β is also based on information from the remote processor 206. Cloud confidence assesses confidence in the current state of the digital map.

[0041] In field 508, the digital map is updated and contains an updated mapped node for the traffic light.

[0042] The confidence coefficient β is used to update the mapped positions of the mapped nodes within the digital map 220. One possible update procedure is shown in Eq. (6): Mi=(1−β)⋅Mi+β⋅Vi

[0043] Even if an observed node is not paired with a mapped node in the digital map 220, a new mapped node can be created within the digital map 220.

[0044] The remote processor 206 manages or tracks a utility value for each mapped node in the digital map 220. When the remote processor 206 receives a new observation from a probe vehicle that pairs with the mapped node, the utility value of the mapped node is increased by a selected utility increment δ. The utility increment δ is a function of several parameters, including the model of the vehicle and the model of the sensor(s) on the vehicle, an observation distance between the probe vehicle and the traffic light, the observation quality of the probe vehicle, and historical data from that vehicle (i.e., does the probe vehicle have a history of detecting false positives?).

[0045] Fig. Diagram 700 illustrates the evolution of the utility value for an illustratively mapped node. Time is plotted along the x-axis and the utility value along the y-axis. The utility value increases at time t1 because a first observation is recorded at the remote processor 206, which verifies the mapped node. At time t2, the utility value increases again because a second observation is recorded, confirming the first. During time intervals in which no observation is received at the remote processor 206, the utility value decreases monotonically. Diagram 700 shows several thresholds: a valid threshold γ1, an invalid threshold γ2, and a distance threshold γ3. If the utility value δ is greater than the valid threshold γ1, the mapped node is considered valid and can be made available to the carrier vehicle for later use.If the utility value falls below the invalid threshold γ2, the mapped node is considered invalid. If the utility value falls below the distance threshold γ3, the mapped node is completely removed from the digital map.

[0046] Fig. Figure 8 shows an illustrative graph of 800 showing an average light position based on a selected number of observations or passes through the intersection point. As the number of observations of the traffic light increases, the mean distance of the traffic light above the ground approaches a selected value.

[0047] While the methods disclosed here are discussed with regard to determining the location or position of a traffic light within an image taken from a vehicle in order to reduce the computation time when determining the state of the traffic light, the methods can be used with regard to any selected object, in particular with regard to reducing the search time for the object within an image taken from a vehicle.

[0048] While the above disclosure has been described with reference to exemplary embodiments, it will be understood by those skilled in the art that various modifications can be made and elements thereof can be replaced by equivalents without departing from its scope. Furthermore, many modifications can be made to adapt a particular situation or material to the teachings of the disclosure without departing from its essential scope. It is therefore intended that the present disclosure is not limited to the individual embodiments disclosed, but includes all embodiments that fall within its scope.

Claims

[1] A method for locating a traffic light (202) on a carrier vehicle (10), comprising: Acquiring image data at an intersection that includes the traffic light (202) on a probe vehicle (204) when the probe vehicle (204) is at the intersection; Identifying the traffic light (202) at the intersection from the image data; Create, on a remote processor (206), an observed node in a digital map (220) corresponding to the traffic light (202); Updating, on the remote processor (206), a mapped position of a mapped node within the digital map (220) based on an observed position of the observed node, wherein the mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map (220); and Locating the traffic light (202) at the carrier vehicle (10) using the mapped node of the digital map (220) when the carrier vehicle (10) is at the intersection. [2] The method according to claim 1, further comprising the use of the mapped node of the digital map (220) to locate the traffic light (202) within image data of the intersection obtained from the carrier vehicle (10). [3] The method according to claim 1, further comprising assigning a utility value to the mapped node in the digital map (220), wherein the utility value increases with a number of observations within a selected time period. [4] The method according to claim 1, which further comprises determining a location of the carrier vehicle (10) from an observation of the traffic light (202) at the carrier vehicle (10) and the mapped position of the mapped node within the digital map (220). [5] The method according to claim 1, further comprising adding a new mapped node to the digital map (220) to represent the traffic light (202) when the observed node associated with the traffic light (202) does not pair with the mapped node. [6] A system for locating a traffic light (202) on a carrier vehicle (10), comprising: at least one probe vehicle (204) configured to receive image data about an intersection that includes the traffic light (202) when the probe vehicle (204) is at the intersection, and to identify the traffic light (202) in the intersection from the image data; and a remote processor (206) configured to: Creating an observed node in a digital map (220) corresponding to the traffic light (202); and Updating a mapped position of a mapped node within the digital map (220) based on an observed position of the observed node, wherein the mapped position of the mapped node and the observed position of the observed node are displayed by three-dimensional coordinates within the digital map (220); and wherein the carrier vehicle (10) uses the mapped node of the digital map (220) to locate the traffic light (202) when the carrier vehicle (10) is at the intersection. [7] The system according to claim 6, wherein the carrier vehicle (10) uses the mapped node of the digital map (220) to locate the traffic light (202) within the image data of the intersection obtained from the carrier vehicle (10). [8] The system according to claim 6, wherein the remote processor (206) is further configured to assign a utility value to the mapped node in the digital map (220), the utility value increasing with a number of observations within a selected time period. [9] The system according to claim 6, wherein the carrier vehicle (10) is configured to determine its location from an observation of the traffic light (202) on the carrier vehicle (10) and the mapped position of the mapped node within the digital map (220). [10] The system according to claim 6, wherein the remote processor (206) is further configured to add a new mapped node to the digital map (220) to represent the traffic light (202) when the observed node associated with the traffic light (202) does not pair with the mapped node.

Images