Method and device for evaluating lane line recognition and vehicle

Abstract

Embodiments of the present application disclose a lane line recognition evaluation method and device and a vehicle. The method comprises: obtaining lane line recognition results of at least two frames and obtaining vehicle speeds corresponding to each frame of lane line recognition result; obtaining target time distances corresponding to each frame of lane line recognition result based on a preset target time and the vehicle speed corresponding to each frame of lane line recognition result; obtaining target sampling points corresponding to each frame of lane line recognition result according to each frame of lane line recognition result and the target time distance corresponding to each frame of lane line recognition result; and obtaining an evaluation result of the lane line recognition result based on the target sampling points corresponding to each frame of lane line recognition result. In this way, the evaluation can be completed without relying on parameters collected by external sensors, thereby reducing the evaluation cost.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and more specifically, to a method, apparatus, and vehicle for lane line recognition evaluation. Background Technology

[0002] Lane detection plays a crucial role in autonomous driving technology. To determine the accuracy of lane recognition results, these results are evaluated. However, in these technologies, in addition to the lane recognition itself, other sensors are required on the vehicle to obtain reference recognition results. These reference results are then compared with the lane recognition results to be evaluated, thus increasing the evaluation cost. Summary of the Invention

[0003] In view of the above problems, this application proposes a lane line recognition evaluation method, device and vehicle to improve the above problems.

[0004] In a first aspect, this application provides a method for evaluating lane line recognition, the method comprising: acquiring at least two frames of lane line recognition results and acquiring the vehicle speed corresponding to each frame of the lane line recognition results; obtaining a target time distance corresponding to each frame of the lane line recognition results based on a preset target time and the vehicle speed corresponding to each frame of the lane line recognition results, wherein the target time distance is the distance the vehicle travels longitudinally under the preset target time; obtaining a target sampling point corresponding to each frame of the lane line recognition results based on each frame of the lane line recognition results and the target time distance corresponding to each frame of the lane line recognition results, wherein the target sampling point is the nearest neighbor sampling point corresponding to the target time distance; and obtaining an evaluation result of the lane line recognition results based on the target sampling point corresponding to each frame of the lane line recognition results.

[0005] Secondly, this application provides a lane line recognition evaluation device, the device comprising: a lane line recognition result acquisition unit, configured to acquire at least two frames of lane line recognition results and acquire the vehicle speed corresponding to each frame of the lane line recognition result; a target time distance acquisition unit, configured to obtain a target time distance corresponding to each frame of the lane line recognition result based on a preset target time and the vehicle speed corresponding to each frame of the lane line recognition result, wherein the target time distance is the distance traveled longitudinally by the vehicle under the preset target time; a target sampling point acquisition unit, configured to obtain a target sampling point corresponding to each frame of the lane line recognition result based on each frame of the lane line recognition result and the target time distance corresponding to each frame of the lane line recognition result, wherein the target sampling point is the nearest neighbor sampling point corresponding to the target time distance; and an evaluation result acquisition unit, configured to obtain an evaluation result of the lane line recognition result based on the target sampling point corresponding to each frame of the lane line recognition result.

[0006] Thirdly, this application provides a vehicle including one or more processors and a memory; one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to perform the methods described above.

[0007] Fourthly, this application provides a computer-readable storage medium storing program code, wherein the above-described method is executed when the program code is run.

[0008] This application provides a lane recognition evaluation method, apparatus, vehicle, and storage medium. After acquiring at least two frames of lane recognition results and the vehicle speed corresponding to each frame of lane recognition results, a target time distance corresponding to each frame of lane recognition results is obtained based on a preset target time and the vehicle speed corresponding to each frame of lane recognition results. Based on each frame of lane recognition results and the target time distance corresponding to each frame of lane recognition results, target sampling points corresponding to each frame of lane recognition results are obtained. Based on the target sampling points corresponding to each frame of lane recognition results, an evaluation result for the lane recognition results is obtained. This method allows for the acquisition of at least two frames of lane recognition results and the corresponding vehicle speed, followed by the acquisition of the target time distance corresponding to each frame of lane recognition results, and the obtaining of the target sampling points corresponding to each frame of lane recognition results. This enables the evaluation of lane recognition results to be obtained based on the target sampling points corresponding to each frame of lane recognition results, without relying on parameters collected by external sensors, thus reducing evaluation costs. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A flowchart of a lane line recognition evaluation method proposed in this application embodiment;

[0011] Figure 2 A schematic diagram of a lane line recognition result proposed in this application is shown;

[0012] Figure 3 A flowchart of an evaluation method for lane line recognition according to another embodiment of this application is shown;

[0013] Figure 4 A schematic diagram of a target sampling point determination method proposed in this application is shown;

[0014] Figure 5 This paper shows a structural block diagram of a lane line recognition evaluation device according to an embodiment of the present application;

[0015] Figure 6 A structural block diagram of a vehicle proposed in this application is shown. Detailed Implementation

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0017] In this application embodiment, the inventors propose a lane line recognition evaluation method, apparatus, and vehicle. After acquiring at least two frames of lane line recognition results and the vehicle speed corresponding to each frame of lane line recognition results, a target time distance corresponding to each frame of lane line recognition results is obtained based on a preset target time and the vehicle speed corresponding to each frame of lane line recognition results. Based on each frame of lane line recognition results and the target time distance corresponding to each frame of lane line recognition results, target sampling points corresponding to each frame of lane line recognition results are obtained. Based on these target sampling points, an evaluation result for the lane line recognition results is obtained. This method allows for the acquisition of at least two frames of lane line recognition results and the vehicle speed corresponding to each frame of lane line recognition results. By obtaining the target time distance corresponding to each frame of lane line recognition results and the target sampling points corresponding to each frame of lane line recognition results, the evaluation result for the lane line recognition results can be obtained based on these target sampling points. This eliminates the need to rely on parameters collected by external sensors, reducing evaluation costs.

[0018] Please see Figure 1 This application provides an evaluation method for lane line recognition, the method comprising:

[0019] S110: Obtain at least two frames of lane line recognition results and obtain the vehicle speed corresponding to each frame of the lane line recognition results.

[0020] Among them, such as Figure 2 As shown, the lane line recognition result can include multiple lane lines, and each lane line includes multiple sampling points and the location information of the multiple sampling points.

[0021] As one approach, vehicles can be equipped with image acquisition devices (cameras, radar, etc.). These devices can collect environmental information such as lane lines, road signs, and traffic lights during vehicle operation at a certain acquisition frequency, thereby obtaining at least two frames of images. Based on these at least two frames of images and a lane line recognition algorithm, at least two frames of lane line recognition results can be obtained.

[0022] In this embodiment of the application, in order to improve the real-time performance of the evaluation, at least two frames of lane line recognition results can refer to two consecutive frames of lane line recognition results.

[0023] In order to determine the evaluation results of lane line recognition results over a period of time while conserving computing resources, at least two frames of lane line recognition results can refer to two frames of lane line recognition results spaced at a preset number of frames apart. For example, the lane line recognition results of the first frame and the lane line recognition results of the fifth frame can be used as at least two frames of lane line recognition results.

[0024] To improve evaluation accuracy, a sliding time window mechanism can be used to obtain lane recognition results from two consecutive sliding time windows, thus obtaining two consecutive sets of lane recognition results. Each set of lane recognition results includes multiple consecutive frames of lane recognition results, and the two consecutive sets of lane recognition results are considered as at least two frames of lane recognition results. For example, the sliding time window can be 3 frames in size. The first set of lane recognition results can include the lane recognition results from frames 1 to 3, and the second set of lane recognition results can include the lane recognition results from frames 2 to 4; or, the first set of lane recognition results can include the lane recognition results from frames 1 to 3, and the second set of lane recognition results can include the lane recognition results from frames 4 to 6.

[0025] Optionally, when at least two frames of lane line recognition results refer to two frames of lane line recognition results with a preset frame interval, the number of frames between the two frames can be determined based on actual needs. The higher the real-time requirement, the fewer the frames can be; the higher the requirement for resource conservation, the more frames can be.

[0026] S120: Based on the preset target time and the vehicle speed corresponding to the lane line recognition result in each frame, the target time distance corresponding to the lane line recognition result in each frame is obtained, where the target time distance is the distance the vehicle travels longitudinally under the preset target time.

[0027] There can be multiple preset target times, and the time interval between adjacent preset target times can be the same. For example, multiple preset target times can be 0 seconds, 0.5 seconds, 1 second, and 1.5 seconds. The time interval between adjacent preset target times can also be different, and they can be sorted in ascending order of preset target time, with the time interval between the later preset target times being larger. For example, multiple preset target times can be 0 seconds, 0.5 seconds, 1.5 seconds, and 3 seconds.

[0028] One approach is to acquire vehicle speed using an accelerometer or similar device while obtaining lane line recognition results for each frame. Based on multiple preset target times and the vehicle speed corresponding to each frame's lane line recognition result, multiple target time distances can be obtained, where each preset target time corresponds to one target time distance.

[0029] For example, for a lane line recognition result, multiple preset target times can be 0 seconds, 0.5 seconds, 1 second, and 1.5 seconds, and the vehicle speed can be V. Then the target time distance can be 0*V, 0.5*V, 1*V, and 1.5*V.

[0030] S130: Based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame, obtain the target sampling point corresponding to the lane line recognition result of each frame, and the target sampling point is the nearest neighbor sampling point corresponding to the target time distance.

[0031] The nearest neighbor sampling point corresponding to the target time distance can be understood as the sampling point closest to the straight line where the target time distance is located. For the same lane line, one target time distance can correspond to one target sampling point.

[0032] One approach is to obtain the target sampling points corresponding to each frame of lane line recognition results based on the lane line recognition results of each frame and the time distance of multiple targets corresponding to each frame of lane line recognition results.

[0033] S140: Based on the target sampling points corresponding to the lane line recognition results in each frame, obtain the evaluation results of the lane line recognition results.

[0034] One approach is to obtain the lateral difference between target sampling points at the same target time distance in lane line recognition results of different frames based on the target sampling points corresponding to each frame of lane line recognition results; and to obtain the evaluation result based on the lateral difference and the time interval between lane line recognition results of different frames, which can characterize the lateral position change rate of the target sampling points.

[0035] The location information of each target sampling point can be represented by a set of coordinates (x, y) in the vehicle coordinate system. The X-axis direction can represent the longitudinal direction of the vehicle, which is equivalent to the extension direction of the lane line. The Y-axis direction can represent the lateral direction of the vehicle. The lateral difference can refer to the distance between two target sampling points in the Y-axis direction.

[0036] Optionally, each frame of lane line recognition results can have a corresponding timestamp, and the time interval between different frames of lane line recognition results can be obtained by using the timestamps corresponding to different frames of lane line recognition results.

[0037] For example, the time interval between lane line recognition results in different frames can be t. For the same lane line in the lane line recognition results of different frames, the target sampling points corresponding to the target time distance 0.5*V can be (x1, y1) and (x2, y2) respectively. Then the lateral difference between the target sampling points can be y2-y1, and the lateral position change rate of the pair of target sampling points can be |(y2-y1) / t|. By taking the average of the lateral position change rates of multiple pairs of target sampling points in the lane line recognition results of different frames, the evaluation result can be obtained.

[0038] In this embodiment of the application, the evaluation results can reflect the changes in the lane line recognition results at different times, so the stability of the lane line recognition results can be obtained through the evaluation results. The smaller the evaluation result, the higher the stability, indicating that the lane line recognition results are more accurate.

[0039] It should be noted that when obtaining two consecutive sets of lane line recognition results, the multiple frames of lane line recognition results in each set can be sorted according to the order of collection time. Then, based on the target sampling points corresponding to every two frames of lane line recognition results in the same position in the two sets, multiple reference evaluation results can be obtained. Each reference evaluation result can be the evaluation result of the corresponding two frames of lane line recognition results. Finally, the average of multiple reference evaluation results is used as the evaluation result of the two consecutive sets of lane line recognition results.

[0040] This embodiment provides a lane recognition evaluation method. After acquiring at least two frames of lane recognition results and the vehicle speed corresponding to each frame of lane recognition results, a target time distance corresponding to each frame of lane recognition results is obtained based on a preset target time and the vehicle speed corresponding to each frame of lane recognition results. Based on each frame of lane recognition results and the target time distance corresponding to each frame of lane recognition results, target sampling points corresponding to each frame of lane recognition results are obtained. Based on these target sampling points, an evaluation result for the lane recognition results is obtained. This method allows for the acquisition of at least two frames of lane recognition results and the corresponding vehicle speed, followed by the acquisition of the target time distance corresponding to each frame of lane recognition results, and the subsequent determination of the target sampling points. This enables the evaluation of lane recognition results to be obtained based on these target sampling points, eliminating the need for parameters collected by external sensors and reducing evaluation costs.

[0041] Please see Figure 3 This application provides an evaluation method for lane line recognition, the method comprising:

[0042] S210: Obtain at least two frames of lane line recognition results and obtain the vehicle speed corresponding to each frame of the lane line recognition results.

[0043] S220: Based on the preset target time and the vehicle speed corresponding to the lane line recognition result in each frame, the target time distance corresponding to the lane line recognition result in each frame is obtained, where the target time distance is the distance the vehicle travels longitudinally under the preset target time.

[0044] S230: Based on the lane line recognition results of each frame, obtain multiple sampling points contained in the target lane line in each frame of the lane line recognition results.

[0045] The target lane line can refer to a lane line of type adjacent to the left lane line (ID can be 0), from the left lane line (ID can be 1), from the right lane line (ID can be 2), or adjacent to the right lane line (ID can be 3), such as... Figure 4 As shown, there are currently 4 target lane lines on the road. The adjacent left lane line refers to the lane line furthest from the vehicle in the left lane of the vehicle's lane. The left lane line refers to the lane line to the left of the vehicle's lane. The right lane line refers to the lane line to the right of the vehicle's lane. The adjacent right lane line refers to the lane line furthest from the vehicle in the right lane of the vehicle's lane.

[0046] One approach is to divide each lane line into multiple lane lines of a fixed length; based on the sampling points included in each lane line, obtain the cluster center of each lane line to obtain multiple cluster centers contained in each lane line; obtain the vehicle position corresponding to each frame of lane line recognition results; based on the distance between the multiple cluster centers contained in each lane line and the vehicle position, obtain the type of each lane line; and take the lane line of type adjacent to the left lane line, or from the left lane line, or from the right lane line, or adjacent to the right lane line as the target lane line to obtain multiple sampling points contained in the target lane line.

[0047] Optionally, the average coordinates of the sampling points included in each lane line in the X and Y directions can be obtained based on the coordinates of the sampling points included in each lane line in the X and Y directions, and the average coordinates of the sampling points included in each lane line in the X and Y directions can be used as the coordinates of the cluster center of each lane line in the X and Y directions.

[0048] For example, please continue reading Figure 4 , Figure 4 The middle and adjacent left lane line can be divided into multiple lane lines. The first lane line can have 3 sampling points, namely (x1, y1), (x2, y2), and (x3, y3). Then the cluster center of the first lane line can be ((x1+x2+x3) / 3, (y1+y2+y3) / 3).

[0049] Optionally, the type corresponding to the multiple cluster centers contained in each lane line can be obtained based on the distance between each cluster center and the vehicle position, and the type that appears most frequently among the multiple cluster centers contained in each lane line can be taken as the type of each lane line.

[0050] The vehicle position can refer to the coordinates of the vehicle in the vehicle coordinate system. The distance between the cluster center and the vehicle position can refer to the difference between the cluster center and the vehicle position in the Y-axis direction, that is, the Y-axis coordinate of the cluster center minus the Y-axis coordinate of the vehicle.

[0051] For example, a lane line can have 5 cluster centers, and the types of the 5 cluster centers can be left-hand lane line, left-hand lane line, adjacent left-hand lane line, left-hand lane line, and left-hand lane line, respectively. Then the type of the lane line is left-hand lane line.

[0052] Optionally, if the distance between the cluster center and the vehicle position is greater than or equal to a first distance threshold and less than a second distance threshold, the cluster center is determined to be an adjacent right lane line; if the distance between the cluster center and the vehicle position is greater than or equal to a second distance threshold and less than a third distance threshold, the cluster center is determined to be a right lane line; if the distance between the cluster center and the vehicle position is greater than or equal to a fourth distance threshold and less than a fifth distance threshold, the cluster center is determined to be a left lane line; if the distance between the cluster center and the vehicle position is greater than or equal to a fifth distance threshold and less than a sixth distance threshold, the cluster center is determined to be an adjacent left lane line. Here, the first, second, and third distance thresholds are all less than 0, and the first distance threshold is less than the second distance threshold, and the second distance threshold is less than the third distance threshold; the fourth, fifth, and sixth distance thresholds are all greater than 0, and the fourth distance threshold is less than the fifth distance threshold, and the fifth distance threshold is less than the sixth distance threshold. These distance thresholds can be determined based on the lane line width.

[0053] In this embodiment of the application, by taking the type that appears most frequently among the multiple cluster centers contained in each lane line as the type of each lane line, misjudgment of lane line type can be avoided, thereby improving the accuracy of the obtained target lane line.

[0054] S240: Based on the target time distance corresponding to the lane line recognition result in each frame, obtain the endpoint corresponding to the target time distance.

[0055] As one approach, after obtaining the target time distance corresponding to the lane line recognition result, the vehicle position can be used as the starting point to obtain the endpoint corresponding to the target time distance based on the starting point and the target time distance.

[0056] For example, Figure 4 Points A, B, C, and D in the diagram can be the endpoints corresponding to the target distance.

[0057] S250: Based on the plurality of sampling points and the endpoint, the target sampling point is obtained.

[0058] One approach is to obtain multiple target straight lines, each of which can be the line where the endpoint of the corresponding target time distance is located, and each target straight line is perpendicular to the longitudinal direction of the vehicle's travel. Based on the distance between each target straight line and multiple sampling points contained in the target lane line in the corresponding lane line recognition result, the sampling point closest to each target straight line is obtained, and the sampling point closest to each target straight line is taken as the target sampling point.

[0059] Optionally, to improve the efficiency of obtaining target sampling points, the sampling point closest to the target line can be obtained through a vertical binary search method.

[0060] For example, such as Figure 4 As shown, the straight lines containing points A, B, C, and D can be straight lines 1, 2, 3, and 4, respectively. By measuring the distances between multiple sampling points and straight lines 1, 2, 3, and 4, we can obtain the sampling points closest to straight lines 1, 2, 3, and 4, respectively. In other words, we can obtain the target sampling points corresponding to straight lines 1, 2, 3, and 4 on each target lane.

[0061] S260: Based on the target sampling points corresponding to the lane line recognition results in each frame, obtain the evaluation results of the lane line recognition results.

[0062] As one approach, when all frames of lane line recognition results contain 4 target lane lines and 4 target time distances, then for each target lane line in each frame of lane line recognition results, there can be 4 target sampling points, which is equivalent to a total of 16 sampling points. By calculating the lateral differences between target sampling points with the same target time distance and the same target lane line in different frames of lane line recognition results, 16 lateral differences can be obtained, thus obtaining 16 reference evaluation results. The average of the 16 reference sampling results is used as the evaluation result of the lane line recognition result.

[0063] This embodiment provides a lane line recognition evaluation method. By acquiring at least two frames of lane line recognition results and the vehicle speed corresponding to each frame of lane line recognition results, the target time distance corresponding to each frame of lane line recognition results is obtained based on a preset target time and the vehicle speed corresponding to each frame of lane line recognition results. This results in the target sampling points corresponding to each frame of lane line recognition results, allowing the evaluation results of lane line recognition results to be obtained based on the target sampling points corresponding to each frame of lane line recognition results. The evaluation can be completed without relying on parameters collected by external sensors, thus reducing the evaluation cost. Furthermore, in this embodiment, by determining the category of lane lines in each frame of lane line recognition results and assigning an ID to each category, the target lane line in each frame of lane line recognition results is obtained. Then, the change trend of the same lane line in the time series can be tracked using the lane line ID. However, the logic for assigning lane line IDs in general lane line recognition tasks is a relatively simple sequential assignment, with no logical connection between IDs before and after in the time series. Additionally, when the recognition task is contaminated by noise or the actual shape of the lane lines changes, the lane line IDs may jump, leading to errors in the matching of target sampling points. Therefore, associating lane lines in the time series based on the relationship between the cluster center of each lane line and the vehicle position can improve the accuracy of the target lane line, thereby avoiding jumps in evaluation results caused by changes in the actual number and shape of lane lines during the evaluation process, and improving evaluation accuracy.

[0064] Please see Figure 5 This application provides a lane line recognition evaluation device 400, the device 400 comprising:

[0065] Lane line recognition result acquisition unit 410 is used to acquire at least two frames of lane line recognition results and acquire the vehicle speed corresponding to each frame of lane line recognition results.

[0066] The target time distance acquisition unit 420 is used to obtain the target time distance corresponding to the lane line recognition result of each frame based on the preset target time and the vehicle speed corresponding to the lane line recognition result of each frame. The target time distance is the distance the vehicle travels longitudinally under the preset target time.

[0067] The target sampling point acquisition unit 430 is used to obtain the target sampling point corresponding to the lane line recognition result of each frame based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame, wherein the target sampling point is the nearest neighbor sampling point corresponding to the target time distance;

[0068] The evaluation result acquisition unit 440 is used to obtain the evaluation result of the lane line recognition result based on the target sampling point corresponding to the lane line recognition result in each frame.

[0069] In one manner, the lane line recognition result acquisition unit 410 is specifically used to acquire two consecutive frames of lane line recognition results; or, acquire two frames of lane line recognition results with a preset number of frames between them; or, acquire two consecutive groups of lane line recognition results, each group of lane line recognition results including multiple consecutive frames of lane line recognition results.

[0070] In one approach, the lane line recognition result includes multiple lane lines, each lane line including multiple sampling points. The target time distance acquisition unit 420 is specifically used to obtain multiple sampling points contained in the target lane line in each frame of the lane line recognition result based on the lane line recognition result of each frame; obtain the endpoint corresponding to the target time distance based on the target time distance corresponding to each frame of the lane line recognition result; and obtain the target sampling point based on the multiple sampling points and the endpoint.

[0071] Alternatively, there are multiple preset target times. The target time distance acquisition unit 420 is specifically used to obtain multiple target time distances corresponding to each frame of the lane line recognition result based on the multiple preset target times and the vehicle speed corresponding to each frame of the lane line recognition result, wherein one preset target time corresponds to one target time distance. The target sampling point acquisition unit 430 is specifically used to obtain the target sampling point corresponding to each frame of the lane line recognition result based on each frame of the lane line recognition result and the multiple target time distances corresponding to each frame of the lane line recognition result.

[0072] In one manner, the target sampling point acquisition unit 430 is specifically used to obtain multiple sampling points contained in the target lane line in each frame of the lane line recognition result based on the lane line recognition result of each frame; to obtain the endpoint corresponding to the target time distance based on the target time distance corresponding to the lane line recognition result of each frame; and to obtain the target sampling point based on the multiple sampling points and the endpoint.

[0073] Optionally, the target sampling point acquisition unit 430 is specifically used to divide each lane line into multiple lane lines according to a fixed length; obtain the cluster center of each lane line based on the sampling points included in each lane line, so as to obtain multiple cluster centers included in each lane line; obtain the vehicle position corresponding to the lane line recognition result of each frame; obtain the type of each lane line based on the distance between the multiple cluster centers included in each lane line and the vehicle position; and take the lane line of type adjacent to the left lane line, or from the left lane line, or from the right lane line, or adjacent to the right lane line as the target lane line, so as to obtain multiple sampling points included in the target lane line.

[0074] Optionally, the target sampling point acquisition unit 430 is specifically used to acquire multiple target straight lines, each target straight line being the straight line where the endpoint of the corresponding target time distance is located, and each target straight line being perpendicular to the longitudinal driving direction of the vehicle; based on the distance between each target straight line and the multiple sampling points contained in the target lane line in the corresponding lane line recognition result, the sampling point closest to each target straight line is obtained, and the sampling point closest to each target straight line is taken as the target sampling point.

[0075] In one approach, the evaluation result acquisition unit 440 is specifically used to obtain the lateral difference between target sampling points at the same target time distance in the lane line recognition results of different frames based on the target sampling points corresponding to the lane line recognition results of each frame; and to obtain the evaluation result based on the lateral difference and the time interval between the lane line recognition results of different frames, wherein the evaluation result characterizes the lateral position change rate of the target sampling points.

[0076] The following will combine Figure 6 This application describes one type of vehicle.

[0077] Please see Figure 6 Based on the aforementioned lane line recognition evaluation method and apparatus, this application embodiment also provides another vehicle 100 capable of executing the aforementioned lane line recognition evaluation method. The vehicle 100 includes a processor 102 and a memory 104. The memory 104 stores a program capable of executing the contents of the aforementioned embodiments, and the backup processor 102 can execute the program stored in the memory 104.

[0078] The processor 102 may include one or more processing cores. The processor 102 connects to various parts within the vehicle 100 using various interfaces and lines, and performs various functions and processes data of the vehicle 100 by running or executing instructions, programs, code sets, or instruction sets stored in the memory 104, and by calling data stored in the memory 104. Optionally, the processor 102 may be implemented using at least one of the following hardware forms: a Neural Network Processing Unit (NPU), a Digital Signal Processing Unit (DSP), a Field-Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA). The processor 102 may integrate one or a combination of several of the following: a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a Neural Network Processing Unit (NPU), and a modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; the NPU handles multimedia data such as video and images; and the modem handles wireless communication. It is understandable that the aforementioned modem may not be integrated into the processor 102, but may be implemented using a separate communication chip.

[0079] The memory 104 may include random access memory (RAM), read-only memory (ROM), and double data rate synchronous dynamic random access memory (DDR). The memory 104 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 104 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described below. The data storage area may also store data created during the use of the vehicle 100 (such as phonebooks, audio and video data, chat log data, etc.).

[0080] This application provides a computer-readable storage medium. The computer-readable storage medium stores program code that can be called by a processor to execute the methods described in the above method embodiments.

[0081] Computer-readable storage media can be electronic storage devices such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, computer-readable storage media include non-transitory computer-readable storage media. The computer-readable storage medium has storage space for program code that performs any of the method steps described above. This program code can be read from or written to one or more computer program products. The program code can be compressed, for example, in a suitable form.

[0082] In summary, the lane recognition evaluation method, apparatus, and vehicle provided in this application, after acquiring at least two frames of lane recognition results and the vehicle speed corresponding to each frame of lane recognition results, obtains the target time distance corresponding to each frame of lane recognition results based on a preset target time and the vehicle speed corresponding to each frame of lane recognition results. Based on each frame of lane recognition results and the target time distance corresponding to each frame of lane recognition results, target sampling points corresponding to each frame of lane recognition results are obtained. Based on the target sampling points corresponding to each frame of lane recognition results, the evaluation result of the lane recognition results is obtained. This method allows for obtaining the target time distance corresponding to each frame of lane recognition results after acquiring at least two frames of lane recognition results and the vehicle speed corresponding to each frame of lane recognition results, thereby obtaining the target sampling points corresponding to each frame of lane recognition results. This enables the evaluation result of the lane recognition results to be obtained based on the target sampling points corresponding to each frame of lane recognition results, without relying on parameters collected by external sensors, thus reducing evaluation costs.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for evaluating lane line recognition, characterized in that, The method includes: Acquire at least two frames of lane line recognition results and acquire the vehicle speed corresponding to each frame of the lane line recognition results; wherein, the lane line recognition results include multiple lane lines, and each lane line includes multiple sampling points and the position information of the multiple sampling points; Based on the preset target time and the vehicle speed corresponding to the lane line recognition result in each frame, the target time distance corresponding to the lane line recognition result in each frame is obtained. The target time distance is the distance the vehicle travels longitudinally under the preset target time. Based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame, the target sampling point corresponding to the lane line recognition result of each frame is obtained, and the target sampling point is the nearest neighbor sampling point corresponding to the target time distance; wherein, the nearest neighbor sampling point corresponding to the target time distance is the sampling point closest to the straight line where the endpoint corresponding to the target time distance is located; Based on the target sampling points corresponding to the lane line recognition results in each frame, the evaluation results of the lane line recognition results are obtained.

2. The method according to claim 1, characterized in that, The step of obtaining the target sampling point corresponding to the lane line recognition result of each frame based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame includes: Based on the lane line recognition results of each frame, multiple sampling points contained in the target lane line in each frame's lane line recognition results are obtained; Based on the target time distance corresponding to the lane line recognition result in each frame, the endpoint corresponding to the target time distance is obtained; The target sampling point is obtained based on the multiple sampling points and the endpoint.

3. The method according to claim 2, characterized in that, The step of obtaining multiple sampling points contained in the target lane line in each frame of the lane line recognition result, based on the lane line recognition result of each frame, includes: Each lane line is divided into multiple lane segments of a fixed length; Based on the sampling points included in each lane line segment, the cluster center of each lane line segment is obtained, so as to obtain multiple cluster centers contained in each lane line. Obtain the vehicle position corresponding to the lane line recognition result in each frame; The type of each lane line is determined based on the distances between the multiple cluster centers contained in each lane line and the vehicle position. The lane line of type adjacent to the left lane line, or from the left lane line, or from the right lane line, or adjacent to the right lane line is used as the target lane line to obtain multiple sampling points contained in the target lane line.

4. The method according to claim 2, characterized in that, There are multiple target time distances, and the target sampling points are obtained based on the multiple sampling points and the endpoint, including: Multiple target straight lines are obtained, each target straight line being the line where the endpoint of the corresponding target time distance is located, and each target straight line is perpendicular to the longitudinal direction of the vehicle's travel; Based on the distance between each target straight line and the multiple sampling points contained in the target lane line in the corresponding lane line recognition result, the sampling point closest to each target straight line is obtained, and the sampling point closest to each target straight line is taken as the target sampling point.

5. The method according to claim 1, characterized in that, The evaluation result of the lane line recognition result is obtained based on the target sampling points corresponding to the lane line recognition result in each frame, including: Based on the target sampling points corresponding to the lane line recognition results in each frame, the lateral differences between target sampling points at the same target time distance in the lane line recognition results of different frames are obtained. The evaluation result is obtained based on the lateral differences and the time interval between lane line recognition results in different frames. The evaluation result characterizes the lateral position change rate of the target sampling point.

6. The method according to claim 1, characterized in that, There are multiple preset target times. The step of obtaining the target time distance corresponding to each frame of lane line recognition results based on the preset target times and the vehicle speed corresponding to each frame of lane line recognition results includes: Based on the multiple preset target times and the vehicle speed corresponding to the lane line recognition result in each frame, multiple target time distances corresponding to the lane line recognition result in each frame are obtained, wherein one preset target time corresponds to one target time distance. The step of obtaining the target sampling point corresponding to the lane line recognition result of each frame based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame includes: Based on the lane line recognition result of each frame and the multiple target time distances corresponding to the lane line recognition result of each frame, the target sampling points corresponding to the lane line recognition result of each frame are obtained.

7. The method according to any one of claims 1-6, characterized in that, The acquisition of at least two frames of lane line recognition results includes: Obtain lane line recognition results for two consecutive frames; or, Obtain two frames of lane line recognition results at a preset frame interval; or... Obtain two consecutive sets of lane line recognition results, each set of lane line recognition results including multiple consecutive frames of lane line recognition results.

8. A lane line recognition evaluation device, characterized in that, The device includes: The lane line recognition result acquisition unit is used to acquire at least two frames of lane line recognition results and acquire the vehicle speed corresponding to each frame of the lane line recognition result; wherein, the lane line recognition result includes multiple lane lines, and each lane line includes multiple sampling points and the position information of multiple sampling points; The target time distance acquisition unit is used to obtain the target time distance corresponding to the lane line recognition result of each frame based on a preset target time and the vehicle speed corresponding to the lane line recognition result of each frame. The target time distance is the distance the vehicle travels longitudinally under the preset target time. The target sampling point acquisition unit is used to obtain the target sampling point corresponding to the lane line recognition result of each frame based on the lane line recognition result of each frame and the target time distance corresponding to the lane line recognition result of each frame. The target sampling point is the nearest neighbor sampling point corresponding to the target time distance. The nearest neighbor sampling point corresponding to the target time distance is the sampling point closest to the straight line where the endpoint corresponding to the target time distance is located. The evaluation result acquisition unit is used to obtain the evaluation result of the lane line recognition result based on the target sampling points corresponding to the lane line recognition result in each frame.

9. A vehicle, characterized in that, Includes one or more processors and memory; One or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program code, wherein the method described in any one of claims 1-7 is executed when the program code is run.

Images