A method, device, vehicle and medium for processing track data

Abstract

The present invention discloses a track data processing method, device, vehicle and medium. According to the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar, it is determined whether the current obstacle is associated with the historical obstacle; if the current obstacle is associated with the historical obstacle, the current track data of the current obstacle is updated according to the current track data of the current obstacle and the previous track data of the historical obstacle. The problem that the autonomous driving vehicle cannot accurately detect the obstacle track data within the perception range around the vehicle with only one radar under low-cost conditions is solved, and the previous obstacle track data detected by a radar and the current obstacle track data are associated and matched while reducing the hardware cost, so as to obtain the effect of stable and accurate obstacle track data.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of autonomous driving vehicles, and in particular, to a method, device, vehicle and medium for processing track data. Background Art

[0002] With the development of intelligent driving technology, the trajectory tracking and processing of autonomous driving vehicles are the key issues that need to be addressed in autonomous driving.

[0003] At present, when existing solutions rely solely on radar to perceive the tracks of obstacles around the vehicle, the track processing method mostly relies on the radar point track clustering algorithm to directly match and associate the track data processed by the radar with the system track data. Since the radar will cause track splitting and ID (Identity document, encoding) jumps for large vehicles and some small vehicles, track tracking errors will occur, which brings great danger to the driving of autonomous vehicles. Summary of the invention

[0004] Embodiments of the present invention provide a track data processing method, device, vehicle and medium to obtain accurate obstacle track data and improve the stability and accuracy of the obstacle track perceived by the autonomous driving vehicle.

[0005] In a first aspect, an embodiment of the present invention provides a method for processing track data, the method comprising:

[0006] Determining whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar;

[0007] If the current obstacle is associated with the historical obstacle, the current track data of the current obstacle is updated according to the current track data of the current obstacle and the previous track data of the historical obstacle.

[0008] In a second aspect, an embodiment of the present invention further provides a track data processing device, the track data processing device comprising:

[0009] An obstacle detection module, used to determine whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar;

[0010] The calculation module updates the current track data of the current obstacle according to the current track data of the current obstacle and the previous track data of the historical obstacle if the current obstacle is associated with the historical obstacle.

[0011] In a third aspect, an embodiment of the present invention further provides an autonomous driving vehicle, the autonomous driving vehicle comprising:

[0012] one or more processors;

[0013] A storage device for storing one or more programs;

[0014] Track data detection module, used to collect obstacle track information;

[0015] When the one or more programs are executed by the one or more processors, the one or more processors implement the track data processing method described in any embodiment of the present invention.

[0016] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the track data processing method described in any embodiment of the present invention.

[0017] The embodiments of the present invention provide a track data processing method, device, vehicle and medium. The radar equipment carried by the autonomous driving vehicle detects and obtains the previous track data of the obstacle (i.e., historical obstacle) in the perception range around the vehicle at the previous moment, and the current track data of the obstacle (i.e., current obstacle) in the perception range around the vehicle at the current moment. The correlation matching algorithm is used to correlate and match the current track data detected by the radar with the previous track data. The matching result is used to determine whether the obstacle detected by the radar at the current moment is correlated with the obstacle detected by the radar at the previous moment. If the matching result proves that the two are correlated, the current track data is updated according to the previous track data to obtain accurate obstacle track data. The present invention solves the problems that autonomous driving vehicles cannot accurately detect obstacle track data within the sensing range around the vehicle with only one radar under low-cost conditions, and that obstacle track data is prone to track data splitting and inaccurate detection when obstacle track data is detected only by radar. The present invention reduces hardware costs while correlating and matching the previous obstacle track data detected by one radar with the current obstacle track data to determine whether the historical obstacle and the current obstacle are the same obstacle, thereby obtaining stable and accurate obstacle track data. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of a method for processing track data provided in Embodiment 1 of the present invention;

[0019] Figure 2 A flow chart of a track data processing method provided in Embodiment 2 of the present invention;

[0020] Figure 3 A vehicle surrounding perception interval division structure diagram provided in the second embodiment of the present invention;

[0021] Figure 4 A flow chart of a method for processing track data provided by Embodiment 3 of the present invention;

[0022] Figure 5 A schematic diagram of the structure of a track data processing device provided in Embodiment 4 of the present invention;

[0023] Figure 6 A schematic diagram of the structure of a vehicle provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0025] Before introducing the embodiments of the present invention, it should be noted that the millimeter wave radar detects the obstacle track data in the sensing range around the vehicle by scanning the feature points in the sensing range by the radar, clustering the scanned feature points into tracks by the point clustering processing algorithm of the millimeter wave radar, and establishing a spatial coordinate system with the vehicle as the origin to obtain the position information of the obstacle in the spatial coordinate system, that is, the obstacle ID. The way the millimeter wave radar processes the obstacle track may cause the feature points of an obstacle detected by the millimeter wave radar to obtain multiple obstacle IDs after clustering processing. Among them, the feature point recognition function of the radar is generally provided by the radar manufacturer. For example, the millimeter wave radar scanned 1000 feature points at the previous moment, and the obstacle ID (i.e., historical obstacle) at the previous moment was obtained by the point clustering processing algorithm as 3; at the current moment, the millimeter wave radar scanned the same obstacle, and the scanned feature points were 1010, of which 400 feature points were processed by the point clustering processing algorithm to obtain the current obstacle ID of 4; the other 610 feature points were processed by the point clustering processing algorithm to obtain another current obstacle ID of 8. That is, the current obstacle ID is split from ID3 at the previous moment into ID4 and ID8 at the current moment. To address this problem, an embodiment of the present invention provides a track data processing method, which associates and matches the previous track data of the historical obstacle with the current track data of the current obstacle, determines the obstacle ID in the split state, and corrects it. The specific implementation method is shown in the following embodiment.

[0026] Embodiment 1

[0027] Figure 1This is a flow chart of a method for processing track data provided in the first embodiment of the present invention. This embodiment can be applied to the situation of how to perceive the track of obstacles, especially to the situation of perceiving the track of obstacles in the vehicle driving scene on the expressway or fast road section. The method can be executed by the track data processing device provided in the embodiment of the present invention, and the device can be implemented in software and / or hardware. The device can be configured in the terminal device / server / vehicle controller, and the method specifically includes:

[0028] S110 : Determine whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar.

[0029] Among them, the previous track data of the historical obstacle is the track data of the obstacle in the sensing interval around the vehicle at the previous moment, and the historical obstacle is the obstacle in the sensing interval around the vehicle at the previous moment. The current track data of the current obstacle is the track data of the obstacle in the sensing interval around the vehicle at the current moment, and the current obstacle is the obstacle in the sensing interval around the vehicle at the current moment. The track data of the obstacle in the sensing interval around the vehicle can be obtained through vehicle radar detection.

[0030] The previous track data of the historical obstacle and the current track data of the current obstacle can be obtained by vehicle radar detection. The vehicle radar can be a radar device mounted on an autonomous vehicle. Preferably, the radar device can be a millimeter wave radar. The millimeter wave radar uses the reflected wave of the detected object to locate and display the echo through the radiated millimeter wave. Its working process is mainly: the radar radiates electromagnetic waves through the radio frequency system to detect and scan the electromagnetic waves of the obstacle, and uses the reflected electromagnetic waves to amplify and analyze the signal. The distance of the obstacle relative to the radar can be calculated, and the left and right distance of the obstacle from the radar can be calculated by combining different azimuth angles. The Doppler effect is used to calculate the moving speed and heading angle of the obstacle. The obstacle is an obstacle that appears in the sensing range around the vehicle during the driving process. For example, it can be a vehicle in the process of moving; the sensing range around the vehicle is the area that can be detected by the millimeter wave radar. The track data includes the moving speed, heading angle, lateral distance from the autonomous driving vehicle, and longitudinal distance from the autonomous driving vehicle.

[0031] Optionally, this step may calculate whether the current obstacle can be associated with the historical obstacle according to the previous track data of the historical obstacle and the current track data of the current obstacle according to an association matching algorithm.

[0032] Specifically, it can be achieved through the following sub-steps:

[0033] S1101, determining a location association threshold of historical obstacles.

[0034] The position association threshold refers to the distance between the vehicle and the obstacle, such as the length distance, width distance and height distance between the vehicle and the obstacle. The position association threshold can be expressed in the form of coordinates.

[0035] Specifically, since most obstacles are not point-shaped objects, but have their own size parameters. When the radar detects the position of an obstacle, it identifies the obstacle as a point. Therefore, before setting the position association threshold, a size threshold of the obstacle is set in advance. Preferably, a size threshold is set as (4m, 1.6m, 2m) based on the length, width and height of a car. Then obtain the previous track data detected by the millimeter-wave radar to obtain the association threshold (X1, X2, X3, X4, X5), where X1 is the obstacle moving speed detected by the millimeter-wave radar, X2 is the heading angle of the obstacle movement, X3 is the lateral distance from the center point of the obstacle to the autonomous driving vehicle, X4 is the longitudinal distance from the center point of the obstacle to the autonomous driving vehicle, and X5 is the height threshold of the obstacle. According to the set size threshold and the previous track data of the obstacle detected by the millimeter-wave radar, an association threshold of the millimeter-wave radar track is set. For example, the previous track data of the detected historical obstacle is: the obstacle's moving speed is 80km / h, the heading angle is 15 degrees, the lateral distance from the autonomous vehicle is 6m, the longitudinal distance from the autonomous vehicle is 20m, and the default obstacle height is 2m; the previous track data of the obstacle is (80,15,6,20,2). According to the previous track data of the obstacle, the obstacle ID is ID3. An obstacle association threshold is set according to the size threshold of the obstacle. A feasible association threshold is: (80,15,6.8,22,2).

[0036] S1102, determining whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle detectable by the radar configured on the vehicle, the current track data of the current obstacle, and a position association threshold.

[0037] Specifically, the millimeter-wave radar detects two obstacles at the current moment. The current track data of the obstacles are: the moving speed of the obstacles is 83km / h, the heading angles are 10 degrees and 12 degrees respectively, the lateral distances from the autonomous driving vehicle are 4m and 6m respectively, the longitudinal distances from the autonomous driving vehicle are 30m and 26m respectively, and the default obstacle height is 2m. Based on this, there are two groups of current track data of the above obstacles, namely (83,10,4,30,2) and (83,12,6,26,2). According to the current track data of the obstacles, the obstacle IDs are ID4 and ID8 respectively.

[0038] According to the position association threshold, the current track data of the obstacle is associated with the previous track data. The association matching can be achieved through an algorithm, such as the Hungarian algorithm. The algorithm can be installed on the vehicle computer, and the vehicle computer can obtain the obstacle track data detected by the radar through the sensor. The current track data and the previous track data are used as algorithm inputs, and the track data are matched and calculated by the Hungarian algorithm. The maximum matching result is compared with the position association threshold to determine whether the current track data has a matching relationship with the previous track data, thereby determining whether the current obstacle track data is associated with the previous obstacle track data, that is, determining whether the current obstacle is associated with the historical obstacle.

[0039] Specifically, if the current track data of the obstacle is associated with the previous track data through the Hungarian algorithm and the maximum matching result is within the association threshold range, then the matching result output is judged to be the association between the current track data and the previous track data, and the historical obstacle and the current obstacle can be associated as the same obstacle.

[0040] S120: If the current obstacle is associated with the historical obstacle, the current track data of the current obstacle is updated according to the current track data of the current obstacle and the previous track data of the historical obstacle.

[0041] Optionally, if the current obstacle is associated with a historical obstacle, the current track data is fused with the previous track data of the historical obstacle associated with the current track data, and the fusion result is used as the updated current track data of the current obstacle.

[0042] Specifically, the current track data of the current obstacle can be two, such as ID4 and ID8. The track data of the historical obstacle associated with the current track data of the previous obstacle is the track data of the historical obstacle that can be associated with the current obstacle as the same obstacle after association matching. The fusion process is to update the current two obstacles according to the previous obstacle track data, such as ID3, that is, to fuse the current track data ID4 and ID8 of the obstacle, fuse ID4 and ID8 into the same ID as the previous obstacle, i.e. ID3, and display the updated data through the vehicle display device.

[0043] If the current obstacle is found to be not associated with the historical obstacle through the association matching algorithm, the current obstacle track data is retained without any processing.

[0044] The technical solution provided in this embodiment provides a track data processing method, device, vehicle and medium. Through the radar equipment carried by the autonomous driving vehicle, the previous track data of the obstacle (i.e., historical obstacle) in the perception range around the vehicle at the previous moment and the current track data of the obstacle (i.e., current obstacle) in the perception range around the vehicle at the current moment are detected and obtained. The correlation matching algorithm is used to correlate the current track data and the previous track data detected by the radar. The matching result is used to determine whether the obstacle detected by the radar at the current moment is associated with the obstacle detected by the radar at the previous moment. If the matching result proves that the two are associated, the current track data is updated according to the previous track data to obtain accurate obstacle track data. The problem that autonomous vehicles cannot accurately detect obstacle track data within the perception range around the vehicle with only one radar under low-cost conditions is solved, and the obstacle track data is prone to track data splitting, resulting in inaccurate detection when obstacle track data is detected only by radar is solved. While reducing hardware costs, the previous obstacle track data detected by a radar is correlated and matched with the current obstacle track data to determine whether the historical obstacle and the current obstacle are the same obstacle, thereby obtaining stable and accurate obstacle track data.

[0045] Optionally, after the current track data of the current obstacle is updated, the driving track of the vehicle may be adjusted according to the updated current track data of the current obstacle.

[0046] Among them, after the current track data of the current obstacle is updated, the accurate track data of the obstacle in the sensing range around the vehicle is obtained, and the current obstacle ID is corrected and updated according to the previous obstacle ID. The vehicle adjusts the vehicle's current driving track according to the updated current obstacle ID. Specifically, the vehicle adjusts the vehicle's current speed and heading angle and other vehicle driving parameters according to the obstacle track data in the vehicle's sensing range. The radar continues to detect the obstacle track data in the sensing range around the vehicle, and the vehicle updates and adjusts the vehicle's driving track in real time according to the radar's detection results of the obstacle track data.

[0047] Embodiment 2

[0048] Figure 2 A flow chart of a method for processing track data provided in the second embodiment of the present invention, Figure 3 This is a diagram of a vehicle surrounding perception interval division structure provided in the second embodiment of the present invention. This embodiment is optimized on the basis of the above embodiment, and provides a preferred embodiment of a method for processing track data based on obstacle data collected by vehicle radar and vehicle camera. Specifically, Figure 2 As shown, the track data processing method provided in this embodiment may include:

[0049] S210: Determine a position association threshold of the historical obstacle according to size parameters of the historical obstacle collected by the vehicle camera.

[0050] In this embodiment, since the vehicle-mounted camera can generally only collect obstacle information in a certain area, and cannot perform obstacle detection without blind spots in the sensing area around the vehicle, when the vehicle collects obstacle track data in its sensing area, it often uses the camera and radar to jointly perform the collection of obstacle track data. Among them, the camera can be mounted on the vehicle, such as Figure 3 As shown, the perception range that can be recognized by the vehicle camera is taken as area one; the perception range from the left and right peripheries of the vehicle body to the camera blind spot is taken as area two; and the perception range of the rear part of the autonomous driving vehicle body is taken as area three.

[0051] Specifically, the vehicle camera can be used to collect obstacle track data and obstacle size parameters in area 1. The size parameters include the length, width and height of the obstacle, which can be expressed as (X, Y, Z), where X is the length of the obstacle, Y is the width of the obstacle, and Z is the height of the obstacle. According to the size data of the obstacle at the previous moment recognized by the camera, the size data of the obstacle is set as the size threshold, and the distance from the center point of the autonomous driving vehicle to the obstacle can be set as the position association threshold. For example, the obstacle size parameters in area 1 collected by the vehicle camera are (6m, 2m, 4m). Then obtain the previous track data of the historical obstacle detected by the vehicle camera, and obtain the association threshold (X1, X2, X3, X4, X5), where X1 is the obstacle moving speed detected by the millimeter wave radar, X2 is the heading angle of the obstacle movement, X3 is the lateral distance from the center point of the obstacle to the autonomous driving vehicle, X4 is the longitudinal distance from the center point of the obstacle to the autonomous driving vehicle, and X5 is the height threshold of the obstacle. According to the set size threshold and the previous track data of the obstacle detected by the millimeter-wave radar, set a millimeter-wave radar track association threshold. For example, the previous track data of the historical obstacle detected is: the obstacle's moving speed is 80km / h, the heading angle is 15 degrees, the lateral distance from the autonomous driving vehicle is 6m, the longitudinal distance from the autonomous driving vehicle is 20m, and the default obstacle height is 4m; the previous track data of the obstacle is (80,15,6,20,4). According to the previous track data of the obstacle, the obstacle ID is ID3. According to the size threshold of the obstacle, set an obstacle association threshold. A feasible position association threshold is: (80,15,7,23,4).

[0052] S220: Determine whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar.

[0053] The previous track data of the historical obstacle is calculated by the above fusion operation. Further, according to the previous track data of the historical obstacle and the current track data of the current obstacle detectable by the radar configured on the vehicle, and the position association threshold, it is determined whether the current obstacle is associated with the historical obstacle.

[0054] Specifically, the radar detects two obstacles at the current moment. The current track data of the obstacles are: the moving speed of the obstacles is 83km / h, the heading angles are 10 degrees and 12 degrees respectively, the lateral distances from the autonomous driving vehicle are 4m and 6m respectively, and the longitudinal distances from the autonomous driving vehicle are 30m and 26m respectively. According to the obstacle size threshold value collected by the vehicle camera, the default obstacle height is 4m. Based on this, there are two groups of current track data of the above obstacles, namely (83,10,4,30,4) or ID4, and (83,12,6,26,4) or ID8.

[0055] According to the position association threshold, the current track data of the obstacle is associated and matched with the previous track data of the historical obstacle. The association matching can be achieved through an algorithm, such as the Hungarian algorithm. The algorithm can be installed on the vehicle computer, and the vehicle computer can obtain the obstacle track data detected by the radar through the sensor. The current track data and the previous track data of the historical obstacle are used as algorithm inputs, and the track data are matched and calculated through the Hungarian algorithm. The maximum matching result is compared with the position association threshold to determine whether the current track data has a matching relationship with the previous track data, thereby determining whether the current obstacle track data is associated with the previous track data of the historical obstacle, that is, determining whether the current obstacle is associated with the historical obstacle.

[0056] S230: If the current obstacle is associated with the historical obstacle, the current track data of the current obstacle is updated according to the current track data of the current obstacle and the previous track data of the historical obstacle.

[0057] In addition, both area 2 and area 3 are camera blind areas. After the position association threshold is determined by the camera, the detection and update of obstacle track data in area 2 and area 3 can be completed only by radar. The specific method is consistent with the method used in the above embodiment 1 and will not be repeated here.

[0058] The technical solution of this embodiment is to first divide the sensing area around the vehicle into three parts, collect the sensing area that the camera can identify, that is, the previous track data and size parameters of the historical obstacles in the previous moment in area 1 through the camera, and calculate the position association threshold through the information collected by the vehicle camera. Then, the previous track data of the historical obstacle at the previous moment is detected by the radar, and the previous track data collected by the camera and the radar are synchronized and fused, and the result after the fusion operation is used as the previous track data of the historical obstacle. Then, the current track data of the obstacle at the current moment is detected by the radar, and the current track data and the previous track data of the historical obstacle are associated and matched by the association matching algorithm. The matching result is used to determine whether the obstacle detected at the current moment is associated with the obstacle detected at the previous moment. If the matching result proves that the two are associated, the current track data is updated according to the previous track data to obtain stable obstacle track data, and the vehicle's driving track can be adjusted according to the obstacle track data. The problem of artificially defining a fixed obstacle size parameter and obtaining a position association threshold based on the fixed size parameter, which causes certain errors in the association matching results of the current obstacle and the historical obstacle track data, is solved. The camera and radar are combined to update the obstacle size parameter and position association threshold in real time, and accurate and flexible obstacle size parameter and position association threshold are obtained, so as to perform more precise association matching on the current track data and the previous track data of the historical obstacle, and obtain a more accurate association matching result.

[0059] In the track data processing method provided in this embodiment, a preferred technical solution is that if the vehicle camera detects the historical obstacle at the previous moment, the previous track data of the historical obstacle is obtained by fusion after time synchronization of the previous track data of the historical obstacle detected by the vehicle radar and the previous track data of the historical obstacle detected by the vehicle camera.

[0060] Specifically, if the track data needs to be time synchronized, a duration threshold can be defined. If the time interval between the time when the radar detects the obstacle track data and the time when the camera detects the obstacle track data is less than the duration threshold, it is determined that the time when the radar and the camera detect the obstacle track data is synchronized.

[0061] For example, the time when the camera collects the track data of the obstacle in area one is defined as the first collection time, and the time when the radar collects the track data of the obstacle in area one is defined as the second collection time. After the camera and the radar collect the previous track data of the obstacle, the first collection time, the second collection time, the previous track data collected by the camera, and the previous track data collected by the radar can be sent to the on-board computer device through the sensor. The on-board computer device calculates the interval between the first collection time and the second collection time. If the interval is within the threshold range, the previous track data collected by the camera and the previous track data collected by the radar are fused to obtain the fused previous track data.

[0062] The fusion operation refers to the fusion calculation of the previous track data collected by the camera and the previous track data collected by the radar through an algorithm, such as the Kalman filter algorithm, to obtain the vehicle's track data fusion information, that is, the previous track data of the historical obstacle after fusion. Exemplarily, the duration threshold can be defined as 0.05s. If the interval between the first acquisition time and the second acquisition time is less than 0.05s, the fusion operation is triggered to obtain the previous track data of the historical obstacle.

[0063] Embodiment 3

[0064] Figure 4 A flow chart of a method for processing track data provided by Embodiment 3 of the present invention. Figure 4 As shown, taking an autonomous driving vehicle as an example, the on-board radar can be a millimeter-wave radar.

[0065] The autonomous driving vehicle uses the on-board computer to determine whether the vehicle is in the ODD (Operational Design Domain) through GPS. In this embodiment, the vehicle ODD refers to the vehicle traveling on a highway. In practical applications, it can also be other ODD environments. If the vehicle is traveling in the ODD, the perception interval around the autonomous driving vehicle is divided. The on-board computer receives the obstacle track data detected by the camera and the millimeter-wave radar, and uses the Kalman filter algorithm to time synchronize the obstacle track data collected by the camera and the obstacle track data collected by the radar. The detected obstacle tracks are processed separately for different perception intervals of the vehicle.

[0066] The vehicle perception interval in which both the millimeter-wave radar and the on-board computer can detect obstacle track data is area 1. The on-board computer receives the previous track data of historical obstacles detected by the camera and the millimeter-wave radar, identifies the size parameters of the obstacle and the position relationship between the obstacle and the vehicle, sets the size threshold and position association threshold of the obstacle based on the above data, and fuses the previous track data detected by the camera and the previous track data detected by the millimeter-wave radar to obtain the fused previous track data. The on-board computer receives the current track data of the current obstacle detected by the millimeter-wave radar, and uses the Hungarian algorithm to associate and match the current track data of the obstacle with the previous track data of the historical obstacle based on the position association threshold. If the current track data has a matching relationship with the previous track data, the on-board computer integrates the current track data detected by the millimeter-wave radar to obtain stable obstacle track data, and performs the next round of tracking calculation on the obstacle track.

[0067] The sensing area on both sides of the vehicle where only the millimeter-wave radar can detect the obstacle track data is area 2. The millimeter-wave radar is used to detect the previous track data, and the size threshold of the obstacle is artificially set according to the conventional size of the vehicle, and the position association threshold is calculated based on the size threshold. The millimeter-wave radar is then used to detect the current track data, and the Hungarian algorithm is used to associate and match the current track data with the previous track data. If the current track data and the previous track data have a matching relationship, the on-board computer will integrate the current track data detected by the millimeter-wave radar to obtain stable obstacle track data, and perform the next round of tracking calculation on the obstacle track.

[0068] The sensing area behind the vehicle is defined as area three, and obstacles detected by the millimeter-wave radar within area three are not processed.

[0069] The technical solution of this embodiment is to divide the perception area around the vehicle traveling in the vehicle ODD environment, and different perception areas use different obstacle detection devices to detect obstacles. For the perception area where both the camera and the radar can detect the obstacle track data, the obstacle track data detected by the camera and the radar are fused, and the fusion result is used as the detected obstacle track data, and the position association threshold can be set according to the obstacle size information collected by the camera; for the perception area where only the radar can detect the obstacle track data, the obstacle track data is detected only by the radar. The problem of inaccurate detection results caused by only detecting the obstacle track in the perception area around the vehicle by radar or camera is solved. The perception area around the vehicle is divided to give full play to the performance advantages of the camera and radar in detecting obstacle track data, so as to obtain more accurate obstacle track data.

[0070] Embodiment 4

[0071] Figure 5 Schematic diagram of the structure of a track data processing device provided in Embodiment 4 of the present invention. This embodiment is applicable to the case where obstacle track data within the sensing range around the vehicle is detected and updated, such as Figure 5 As shown, the track data processing device includes: an obstacle detection module 410 and a calculation module 420.

[0072] The obstacle detection module 410 is used to determine whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar;

[0073] The calculation module 420 is used to update the current track data of the current obstacle according to the current track data of the current obstacle and the previous track data of the historical obstacle.

[0074] The technical solution provided in this embodiment detects and obtains the previous track data of obstacles in the sensing range around the vehicle at the previous moment, and the current track data of obstacles in the sensing range around the vehicle at the current moment, through the radar equipment carried by the autonomous driving vehicle. Obstacle track data is obtained according to the vehicle size and obstacle track data, and the position association threshold of the obstacle is determined by the previous track data of the obstacle at the previous moment. The association matching algorithm is adopted to associate and match the current track data detected by the radar with the previous track data, and the matching result is used to determine whether the obstacle detected by the radar at the current moment is associated with the obstacle detected by the radar at the previous moment. If the matching result proves that the two are associated, the current track data is updated according to the previous track data to obtain stable obstacle track data, and the vehicle's driving track can be adjusted according to the obstacle track data. The problem that the autonomous driving vehicle cannot accurately detect the obstacle track data in the sensing range around the vehicle by only one radar under low-cost conditions is solved, and the previous obstacle track data detected by one radar and the current obstacle track data are associated and matched while reducing the hardware cost to obtain the effect of stable and accurate obstacle track data.

[0075] The obstacle detection module 410 may be specifically used for:

[0076] Determine the location correlation threshold of the historical obstacle;

[0077] Whether the current obstacle is associated with the historical obstacle is determined based on the previous track data of the historical obstacle, the current track data of the current obstacle detected by the vehicle radar, and the position association threshold.

[0078] Among them, determining the location association threshold of the historical obstacle includes:

[0079] According to the size parameters of the historical obstacles collected by the vehicle camera, the position association threshold of the historical obstacles is determined.

[0080] Furthermore, the calculation module 420 may be specifically used for:

[0081] The current track data of the current obstacle and the previous track data of the historical obstacle associated with the current track data are fused, and the processing result is used as the updated current track data of the current obstacle.

[0082] If the vehicle camera detects a historical obstacle at the previous moment, the previous track data of the historical obstacle is obtained by merging the previous track data of the historical obstacle detected by the vehicle radar at the previous moment and the previous track data of the historical obstacle detected by the vehicle camera after time synchronization.

[0083] Among them, after updating the current track data of the current obstacle, it also includes:

[0084] The driving track of the vehicle is adjusted according to the updated current track data of the current obstacle.

[0085] Exemplarily, the above-mentioned device also includes: a camera, a sensor, an on-board computer and a radar.

[0086] The vehicle positioning device provided in this embodiment can be applied to the track data processing method provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0087] Embodiment 5

[0088] Figure 6 This is a schematic diagram of the structure of a vehicle provided in Embodiment 5 of the present invention. Figure 6 As shown, the vehicle includes a processor 510, a memory 520 and a detector 530; the number of processors 510 in the vehicle can be one or more. Figure 6 A processor 510 is taken as an example; the processor 510, the memory 520 and the detector 530 in the vehicle can be connected via a bus or other means. Figure 6 The example of connecting through bus is taken in the following.

[0089] The memory 520, as a computer-readable storage medium, can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the track data processing method in the embodiment of the present invention (for example, the obstacle detection module 410 and the calculation module 420 in the track data processing device). The processor 510 executes various functional applications and data processing of the vehicle by running the software programs, instructions and modules stored in the memory 520, that is, realizes the positioning method of the above-mentioned autonomous driving vehicle.

[0090] The memory 520 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 520 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 520 may further include a memory remotely arranged relative to the processor 510, and these remote memories may be connected to the vehicle via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0091] The detector 530 may be used to obtain obstacle track data within the vehicle sensing range and track the track of the obstacle.

[0092] The vehicle provided in this embodiment can be applied to the track data processing method provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0093] Embodiment 6

[0094] Embodiment 6 of the present invention further provides a storage medium containing computer executable instructions, wherein the computer executable instructions are used to execute a track data processing method when executed by a computer processor, the method comprising:

[0095] Determining whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar;

[0096] If the current obstacle is associated with the historical obstacle, the current track data of the current obstacle is updated according to the current track data of the current obstacle and the previous track data of the historical obstacle.

[0097] Of course, the computer executable instructions of a storage medium including computer executable instructions provided by an embodiment of the present invention are not limited to the operations of the method described above, but can also execute related operations in the track data processing method provided by any embodiment of the present invention.

[0098] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0099] It is worth noting that in the embodiment of the above-mentioned track data processing method, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0100] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for processing track data, characterized in that: include: Determining whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar; If the current obstacle is associated with the historical obstacle, updating the current track data of the current obstacle according to the current track data of the current obstacle and the previous track data of the historical obstacle; Wherein, determining whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar includes: Determine the position association threshold of the historical obstacle according to the previous track data of the historical obstacle; wherein the previous track data of the historical obstacle includes the obstacle movement speed, the obstacle movement heading angle, the lateral distance between the center point of the obstacle and the autonomous driving vehicle, the longitudinal distance between the center point of the obstacle and the autonomous driving vehicle, and the height of the obstacle; Determining whether the current obstacle is associated with the historical obstacle according to the previous track data of the historical obstacle, the current track data of the current obstacle detected by the vehicle radar, and the position association threshold; Wherein, determining the position association threshold of the historical obstacle according to the previous track data of the historical obstacle includes: For area 1, a position association threshold of the historical obstacle is determined according to the previous track data of the historical obstacle and the size parameter of the historical obstacle collected by the vehicle camera; wherein area 1 is a vehicle perception interval in which both the vehicle radar and the vehicle camera can detect the obstacle track data; For area two, the position association threshold is determined based on the previous track data of the historical obstacle and the size threshold of the obstacle set according to the conventional size of the obstacle vehicle; wherein area two is the perception interval on both sides of the vehicle where only the vehicle radar detects the obstacle track data.

2. The method according to claim 1, characterized in that Updating the current track data of the current obstacle according to the current track data of the current obstacle and the previous track data of the historical obstacle includes: The current track data of the current obstacle and the previous track data of the historical obstacle associated with the current track data are fused, and the processing result is used as the updated current track data of the current obstacle.

3. The method according to claim 1, characterized in that include: If the vehicle camera detects the historical obstacle at the previous moment, the previous track data of the historical obstacle is obtained by merging the previous track data of the historical obstacle detected by the vehicle radar at the previous moment and the previous track data of the historical obstacle detected by the vehicle camera after time synchronization.

4. The method according to claim 1, characterized in that: Updating the current track data of the current obstacle also includes: The driving track of the vehicle is adjusted according to the updated current track data of the current obstacle.

5. A track data processing device, characterized in that: include: An obstacle detection module, used to determine whether the current obstacle is associated with the historical obstacle based on the previous track data of the historical obstacle and the current track data of the current obstacle detected by the vehicle radar; a calculation module, which updates the current track data of the current obstacle according to the current track data of the current obstacle and the previous track data of the historical obstacle if the current obstacle is associated with the historical obstacle; Wherein, the obstacle detection module is specifically used for: Determine the position association threshold of the historical obstacle according to the previous track data of the historical obstacle; wherein the previous track data of the historical obstacle includes the obstacle movement speed, the obstacle movement heading angle, the lateral distance between the center point of the obstacle and the autonomous driving vehicle, the longitudinal distance between the center point of the obstacle and the autonomous driving vehicle, and the height of the obstacle; Determining whether the current obstacle is associated with the historical obstacle according to the previous track data of the historical obstacle, the current track data of the current obstacle detected by the vehicle radar, and the position association threshold; Wherein, determining the position association threshold of the historical obstacle according to the previous track data of the historical obstacle includes: For area 1, a position association threshold of the historical obstacle is determined according to the previous track data of the historical obstacle and the size parameter of the historical obstacle collected by the vehicle camera; wherein area 1 is a vehicle perception interval in which both the vehicle radar and the vehicle camera can detect the obstacle track data; For area two, the position association threshold is determined based on the previous track data of the historical obstacle and the size threshold of the obstacle set according to the conventional size of the obstacle vehicle; wherein area two is the perception interval on both sides of the vehicle where only the vehicle radar detects the obstacle track data.

6. An autonomous driving vehicle, characterized in that: The vehicle comprises: one or more processors; A storage device for storing one or more programs; Track data detection module, used to collect obstacle track information; When the one or more programs are executed by the one or more processors, the one or more processors implement the track data processing method as described in any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the track data processing method as described in any one of claims 1 to 4 is implemented.

Images