Lane recognition evaluation method, device, electronic device and storage medium

Abstract

The present application discloses a lane line recognition evaluation method, device, electronic device, and storage medium, relating to the field of vehicle technology. The method comprises: obtaining continuous multi-frame lane line recognition results; obtaining lane line parameters corresponding to each frame based on the lane line recognition results of each frame; obtaining a difference value of lane line parameters of adjacent frames based on the lane line parameters corresponding to each frame; and determining an evaluation result of the lane line recognition performance of the target vehicle based on the difference value. In this way, the accuracy and stability of the lane line recognition results are evaluated directly based on the difference value between the lane line parameters of adjacent frames identified. The evaluation can be completed without relying on parameters collected by external reference sensors that require post-processing and manual correction, resulting in high evaluation efficiency and low evaluation cost.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a lane recognition evaluation method, device, electronic device, and storage medium. Background Art

[0002] Lane detection plays a particularly important role in autonomous driving technology. In related technologies, lane recognition results are evaluated to determine their accuracy. However, in related technologies, in addition to the lane recognition that originally needs to be evaluated, other reference sensors need to be configured on the vehicle to obtain reference recognition results. The reference recognition results need to be post-processed and manually corrected, and then the reference recognition results after post-processing and manual correction are compared with the lane recognition results that need to be evaluated. Therefore, the evaluation efficiency is low. Summary of the Invention

[0003] This application proposes a lane line recognition evaluation method, device, electronic device and storage medium to improve the evaluation efficiency of lane line recognition.

[0004] In a first aspect, an embodiment of the present application provides a lane line recognition evaluation method, the method comprising: obtaining continuous multi-frame lane line recognition results; obtaining lane line parameters corresponding to each frame based on the lane line recognition results of each frame; obtaining a difference degree value of lane line parameters of adjacent frames based on the lane line parameters corresponding to each frame; and determining an evaluation result of the lane line recognition performance of the target vehicle based on the difference degree value.

[0005] In an optional embodiment, determining the evaluation result of the lane line recognition performance of the target vehicle based on the difference degree value includes: obtaining the number of the difference degree values ​​greater than a first threshold as a first number, and the total number of the difference degree values ​​as a second number; based on the ratio of the first number to the second number, determining the evaluation result of the lane line recognition stability of the target vehicle, the lane line recognition stability is negatively correlated with the ratio.

[0006] In an optional embodiment, the lane line parameter includes multiple sub-parameters, and the difference degree value includes a difference value corresponding to each sub-parameter; obtaining the number of the difference degree values ​​greater than a first threshold as a first number includes: obtaining the number of difference degree values ​​in which there are target difference values ​​greater than a target parameter threshold as the first number, the target difference value is the difference value corresponding to any sub-parameter, and the target parameter threshold corresponds to the sub-parameter corresponding to the target difference value.

[0007] In an optional embodiment, obtaining the degree of difference of lane line parameters of adjacent frames based on the lane line parameters corresponding to each frame includes: obtaining the absolute value of the difference between each sub-parameter of adjacent frames as the difference value corresponding to each sub-parameter.

[0008] In an optional embodiment, after determining the evaluation result of the lane line recognition stability of the target vehicle based on the ratio of the first number to the second number, the method further includes: if the ratio is greater than the stability threshold, outputting a prompt message, wherein the prompt message is used to prompt that the lane line recognition stability of the target vehicle does not meet the corresponding stability requirements.

[0009] In an optional embodiment, the lane line parameters include the angle of the lane line, and the recognition result includes at least a lane line image, wherein the lane line image contains the recognized lane line; the lane line parameters corresponding to each frame are obtained based on the lane line recognition result of each frame, including: for the lane line contained in each lane line image, obtaining the target part of the lane line whose distance from the target vehicle is less than a first distance; extracting the lane line center point of the target part to obtain multiple lane line center points; fitting the multiple lane line center points to obtain a center point straight line; obtaining the angle between the center point straight line and the target straight line as the angle of the lane line, and the target straight line is the straight line of the travel direction of the target vehicle when the recognition result is recognized.

[0010] In an optional embodiment, obtaining continuous multi-frame lane line recognition results includes: obtaining a road section with a curvature less than a preset curvature from a target map as the target road section; controlling the target vehicle to travel on the target road section, and performing lane line recognition on the target road during driving to obtain continuous multi-frame lane line recognition results.

[0011] In the second aspect, an embodiment of the present application provides a lane line recognition evaluation device, which includes: a recognition result acquisition module for obtaining continuous multi-frame lane line recognition results; a parameter acquisition module for obtaining the lane line parameters corresponding to each frame based on the lane line recognition results of each frame; a difference degree acquisition module for obtaining the difference degree values ​​of the lane line parameters of adjacent frames based on the lane line parameters corresponding to each frame; and an evaluation module for determining the evaluation results of the lane line recognition performance of the target vehicle based on the difference degree values.

[0012] In a third aspect, an embodiment of the present application provides an electronic device comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the above-mentioned method.

[0013] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which program code is stored. The program code can be called by a processor to execute the above method.

[0014] In the solution provided by this application, continuous lane line recognition results are obtained for multiple frames; based on the lane line recognition results for each frame, the lane line parameters corresponding to each frame are obtained; based on the lane line parameters corresponding to each frame, the difference value of the lane line parameters of adjacent frames is obtained; and based on the difference value, the evaluation result of the lane line recognition performance of the target vehicle is determined. In this way, the accuracy and stability of the lane line recognition results can be evaluated directly based on the difference value between the identified lane line parameters of adjacent frames. The evaluation can be completed without relying on parameters collected by external reference sensors, thereby reducing the evaluation cost. Moreover, after obtaining the difference value between the lane line parameters of adjacent frames, no post-processing or manual correction is required. Instead, the lane line recognition performance of the target vehicle is evaluated directly based on the difference value, greatly improving the evaluation efficiency of lane line recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 A flow chart of a lane line recognition evaluation method provided in an embodiment of the present application is shown.

[0017] Figure 2 A flow chart of a lane recognition evaluation method provided in another embodiment of the present application is shown.

[0018] Figure 3 This is a block diagram of a lane recognition evaluation device provided according to an embodiment of the present application.

[0019] Figure 4 It is a block diagram of an electronic device for executing the evaluation method for lane line recognition according to an embodiment of the present application.

[0020] Figure 5 It is a storage unit for storing or carrying program codes for implementing the lane line recognition evaluation method according to the embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0022] It should be noted that in some of the processes described in the specification, claims and the above-mentioned figures of this application, multiple operations that appear in a specific order are included, and these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The sequence numbers of the operations, such as S110, S120, etc., are merely used to distinguish between different operations, and the sequence numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. Also, the terms "first", "second", etc. in the specification, claims and the above-mentioned figures of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or server that includes a series of steps or sub-modules is not necessarily limited to those steps or sub-modules explicitly listed, but may include other steps or sub-modules not explicitly listed or inherent to such process, method, product, or device.

[0023] Research on autonomous driving technology has become increasingly active in recent years. However, fully understanding the surrounding environment remains a significant challenge for autonomous vehicles. Among these environmental perception tasks, lane detection plays a particularly important role in traffic scene recognition. By providing fundamental road information, such as lane structure and the vehicle's position relative to the lanes, accurate lane detection ensures the safe positioning of autonomous vehicles within the main lane.

[0024] In related technologies, evaluation schemes based on reference values ​​evaluate lane line perception results from the perspective of "accuracy". The acquisition of such reference values ​​often relies on expensive measurement equipment and requires post-processing or manual correction, which limits the real-time performance and throughput of the evaluation task.

[0025] The inventors have proposed a lane recognition evaluation method, device, electronic device, and storage medium. These methods directly compare adjacent frames of lane recognition results, evaluating lane recognition results from the perspectives of stability and accuracy, thereby achieving the advantages of low cost and high efficiency. The lane recognition evaluation method provided in the embodiments of this application is described in detail below.

[0026] Please refer to Figure 1 , Figure 1 A flow chart of a lane line recognition evaluation method provided in an embodiment of the present application is provided below. Figure 1 The lane line recognition evaluation method provided in the embodiment of the present application is described in detail. The lane line recognition evaluation method may include the following steps:

[0027] Step S110: Obtaining continuous multi-frame lane line recognition results.

[0028] In this embodiment, the lane recognition performance of any vehicle can be evaluated. For ease of description, the target vehicle is used as the vehicle whose lane recognition performance is to be evaluated. The target vehicle can be a vehicle with autonomous driving capabilities and can be pre-configured with target sensors to collect driving environment data. Lane markings on the road can then be identified based on the driving environment data.

[0029] The target sensor is generally an image sensor, and the corresponding driving environment data collected is also image data. Of course, the target sensor can also be other sensors such as radar, and the corresponding driving environment data collected can be other data such as radar data. This embodiment does not limit this.

[0030] It can be understood that the target sensor has its own data collection frequency. Therefore, when collecting data on the target road section, the target sensor collects data according to its own collection frequency to obtain multiple frames of driving environment data; and for each frame of driving environment data, lane line recognition is performed to obtain the recognition result of the lane line corresponding to each frame of driving environment data, and finally the above-mentioned continuous multiple frames of lane line recognition results can be obtained. The recognition result of each frame of lane line can be regarded as identifying the position of the lane line in each frame of driving environment data; for example, the target sensor is an image sensor, and the driving environment data is multiple frames of driving environment images. In this case, the recognition result is to identify which part of each frame of driving environment image is the lane line. Of course, this embodiment is not limited to the vision-based lane line recognition method, and any other method that can be used for lane line recognition is within the scope of protection of this application.

[0031] Of course, the lane line recognition process can be performed by the target vehicle; or the driving environment data can be uploaded to the corresponding target server and executed by the target server. This embodiment does not limit this.

[0032] In some implementations, lane recognition results are evaluated from a stability perspective. Therefore, a road section with a curvature less than a preset curvature can be obtained from the target map and used as the target section. The target vehicle is then controlled to travel along the target section, and lane recognition is performed on the target road during driving, resulting in multiple frames of continuous lane recognition results. Obviously, due to the low curvature of the target section, the lane lines within the target section will also vary slightly, and will not include lane lines in scenes such as intersections where lane line states fluctuate dramatically. This facilitates the subsequent evaluation of lane recognition.

[0033] In this mode, the user automatically selects a road segment based on the target map, and then automatically identifies road segments with a curvature less than a preset value as target segments. This allows all road segments with a curvature less than a preset value to be automatically selected with one click, eliminating the need for the user to click each one individually. This simplifies the operation and reduces manual work for the evaluator.

[0034] In this mode, in response to a user's segment selection operation based on the target map, if the curvature of the segment selected by the segment selection operation is determined to be less than a preset curvature, the selected segment will be used as the target segment; if the curvature of the segment selected by the segment selection operation is determined to be greater than or equal to the preset curvature, a segment prompt message will be output to inform the evaluator that the curvature of the segment does not meet the evaluation conditions; if a confirmation instruction based on the segment prompt message is received, the selected segment will be used as the target segment. A manual segment selection function is provided for evaluators, allowing them to select a segment that matches the evaluation requirements as the target segment according to changes in the evaluation requirements, thereby meeting different evaluation requirements.

[0035] Step S120: Obtain lane line parameters corresponding to each frame based on the lane line recognition results of each frame. Lane line parameters are identified for each frame recognition result to obtain lane line parameters corresponding to the recognition results of each frame.

[0036] Furthermore, after obtaining the lane line recognition results of each frame, the lane line parameters of the lane lines in the lane line recognition results of each frame can be identified to obtain the lane line parameters corresponding to the lane line recognition results of each frame, that is, the lane line parameters corresponding to each frame mentioned above. Among them, the lane line parameters can be parameters such as the curvature of the lane line, the length of the lane line, the angle of the lane line, or the intercept between the target vehicle and the lane line, and this embodiment does not limit this. Among them, the curvature refers to the curvature of the lane line, and the curvature is used to indicate the deflection direction and degree of deflection of the lane line. The angle of the lane line is the angle between the proximal lane line and the direction of travel of the target vehicle, which is used to indicate the relative relationship between the target vehicle and the direction of travel and the direction of the lane line. Among them, different lane line parameters are identified using different parameter recognition algorithms.

[0037] Step S130: Obtaining the difference degree values ​​of the lane line parameters of adjacent frames according to the lane line parameters corresponding to each frame.

[0038] Step S140: Determine the evaluation result of the lane recognition performance of the target vehicle based on the difference degree value.

[0039] It is understandable that since the target road section contains lane lines that do not change drastically, that is, the shape and direction of the lane lines often do not change significantly, and since the acquisition frequency of the target sensor is relatively high, generally greater than 10 frames / second, the distance traveled by the target vehicle between the lane line recognition results of adjacent frames is very short, which means that the changes in the lane line parameters corresponding to the lane line recognition results of adjacent frames are also very small; if the changes in the lane line parameters corresponding to the lane line recognition results of adjacent frames are very large, it indicates that there are false detections in the lane line recognition results of adjacent frames.

[0040] Therefore, after obtaining the lane line parameters corresponding to each frame, the difference degree values ​​between the lane line parameters of adjacent frames can be obtained to obtain the difference degree values ​​corresponding to adjacent frames. Furthermore, the difference degree values ​​corresponding to adjacent frames can be used to determine whether there is false detection in adjacent frames, that is, to evaluate the accuracy of lane line recognition of the target vehicle; at the same time, based on adjacent frames, the stability of lane line recognition of the target vehicle can be evaluated. Among them, the difference degree value between lane line parameters can be the difference between lane line parameters, or the difference between the normalized values ​​of lane line parameters. The specific method of measuring the difference degree between lane lines is not limited. The evaluation result can be a specific evaluation score, and the lane line recognition performance is positively correlated with the evaluation score; of course, the evaluation result can also have a text description of the lane line recognition performance, so that the evaluator can more intuitively understand the performance of lane line recognition.

[0041] Specifically, if the difference values ​​of adjacent frames are within the first fluctuation range, it is determined that there is no false detection in the adjacent frames; if the difference values ​​corresponding to adjacent frames are within the first fluctuation range, it is determined that there is false detection in the adjacent frames.

[0042] For example, the first fluctuation range is [-1, 1]. If the difference degree value corresponding to the adjacent frames consisting of the first frame and the second frame is 0.5, it is determined that there is no false detection in the adjacent frames; if the difference degree value corresponding to the adjacent frames is -2, it is determined that there is a false detection in the adjacent frames.

[0043] It is understood that the aforementioned continuous multi-frame lane marking recognition results can be obtained while the target vehicle is traveling on the target road section. In one possible implementation, if the lane marking recognition algorithm has good accuracy and stability, the difference between multiple lane marking recognition results at the same location should be very small or close to zero. Therefore, the aforementioned continuous multi-frame lane marking recognition results can also be obtained when the target vehicle is parked at a certain location on the target road section.

[0044] In this embodiment, the accuracy and stability of the lane line recognition results can be evaluated directly based on the difference value between the lane line parameters of adjacent frames. The evaluation can be completed without relying on the parameters collected by external reference sensors, thereby reducing the evaluation cost. Moreover, after obtaining the difference value between the lane line parameters of adjacent frames, there is no need for post-processing or manual correction. Instead, the lane line recognition performance of the target vehicle is evaluated directly based on the difference value, which greatly improves the evaluation efficiency of lane line recognition.

[0045] Please refer to Figure 2 , Figure 2 A flow chart of a lane line recognition evaluation method provided in another embodiment of the present application is shown below. Figure 2 The lane line recognition evaluation method provided in the embodiment of the present application is described in detail. The lane line recognition evaluation method may include the following steps:

[0046] Step S210: Obtaining continuous multi-frame lane line recognition results.

[0047] In this embodiment, the specific implementation of step S210 can refer to the content of the above embodiments and will not be repeated here.

[0048] Step S220: Obtain lane line parameters corresponding to each frame based on the lane line recognition results of each frame.

[0049] In some embodiments, the lane line parameters may include the angle of the lane line, and the recognition result includes at least a lane line image, wherein the lane line image includes the recognized lane line. Specifically, for each lane line included in the lane line image, a target portion of the lane line whose distance from the target vehicle is less than a first distance is obtained; the lane line center point of the target portion is extracted to obtain a plurality of lane line center points; the plurality of lane line center points are fitted to obtain a center point straight line; the angle between the center point straight line and a target straight line is obtained as the lane line angle, wherein the target straight line is the straight line in which the target vehicle is traveling when the recognition result is obtained; and the difference degree values ​​between the lane line parameters of adjacent frames are obtained to obtain the difference degree values ​​of the adjacent frames.

[0050] In this method, multiple lane center points can be fitted using the least squares method. Other fitting algorithms can also be used, without limitation. The lane angle ranges from -90° to 90°, with no restrictions on which direction is considered positive. The first distance can be a preset length, such as 30 meters or 40 meters. This length can also be adjusted based on actual application, without limitation.

[0051] In this embodiment, the lane line parameters may be one or more types, and different lane line parameters are calculated using different parameter algorithms. For example, the curvature of the lane line is calculated using a preset curvature algorithm, which is not limited in this embodiment.

[0052] Step S230: Obtaining the difference degree values ​​of the lane line parameters of adjacent frames according to the lane line parameters corresponding to each frame.

[0053] Optionally, the lane line parameter is the curvature cur of the lane line. Taking the adjacent frames as the first frame and the second frame as an example, the curvatures corresponding to the two frames can be expressed as cur_1 and cur_2 respectively, and |cur_1-cur_2| can be obtained as the difference degree value J_cur[1] corresponding to the adjacent frames.

[0054] Optionally, the lane line parameter is the lane line angle ang. Taking the adjacent frames as the first frame and the second frame as an example, the angles corresponding to the two frames can be expressed as ang_1 and ang_2 respectively, and |ang_1-ang_2| can be obtained as the difference degree value J_ang[1] corresponding to the adjacent frames.

[0055] Optionally, the lane line parameter is the length len of the lane line. Taking the adjacent frames as the first frame and the second frame as an example, the lengths corresponding to the two frames can be expressed as len_1 and len_2 respectively, and |len_1-len_2| can be obtained as the difference degree value J_len[1] corresponding to the adjacent frames.

[0056] In other embodiments, if the lane parameter includes multiple sub-parameters, the difference value includes the difference value corresponding to each sub-parameter. Based on this, the absolute value of the difference between each sub-parameter can be obtained for each pair of adjacent frames as the difference value corresponding to each sub-parameter.

[0057] Alternatively, if the multiple seed parameters include lane curvature, lane angle, and lane length, the corresponding difference values ​​include a difference value corresponding to the curvature, a difference value corresponding to the angle, and a difference value corresponding to the length. Still taking the aforementioned example where the adjacent frames are the first and second frames, the corresponding difference values ​​include J_cur[1], J_ang[1], and J_len[1].

[0058] Step S240: Obtain the number of difference degree values ​​greater than the first threshold as the first number, and the total number of difference degree values ​​as the second number. Obtain the number of recognition result pairs whose difference degree values ​​are greater than the first threshold as the first number, and the total number of recognition result pairs as the second number.

[0059] In this embodiment, a first threshold can be preset. The first threshold can be understood as the maximum permissible threshold for the lane parameter jump amplitude. In other words, if the difference value is greater than the first threshold, it indicates that the lane parameter jump amplitude has exceeded the maximum permissible threshold, and it can be determined that the lane recognition result of at least one frame in the adjacent frames corresponding to the difference value has a false positive.

[0060] Furthermore, the number of difference values ​​greater than a first threshold can be obtained as a first number. Adjacent frames corresponding to difference values ​​greater than the first threshold can be considered abnormal jump frame pairs, and the first number is the number of abnormal jump frame pairs. For example, 10 frames of lane line recognition results can constitute 9 pairs of adjacent frames, i.e., there will be 9 corresponding difference values, and the 9 difference values ​​correspond one-to-one to the 9 pairs of adjacent frames. If 3 of the 9 difference values ​​are greater than the first threshold, the first number is 3.

[0061] In other embodiments, if the lane line parameter includes multiple sub-parameters, the aforementioned difference degree value includes the difference value corresponding to each sub-parameter. In this case, the number of difference degree values ​​for which a target difference value is greater than a target parameter threshold is obtained as the first number, where the target difference value is the difference value corresponding to any sub-parameter, and the target parameter threshold corresponds to the sub-parameter corresponding to the target difference value.

[0062] For example, taking the curvature, angle, and length of lane lines as examples of multiple sub-parameters, the difference values ​​corresponding to the curvature, angle, and length of any pair of adjacent frames can be expressed as J_cur[n], J_ang[n], and J_len[n], respectively. The parameter thresholds corresponding to the curvature are expressed as T_cur, T_ang, and T_len, respectively. The three difference values ​​of any pair of adjacent frames meet at least one of the following conditions:

[0063] J_cur[n]>T_cur

[0064] J_ang[n]>T_ang

[0065] J_len[n]>T_len

[0066] It can be determined that the adjacent frames are the aforementioned abnormal jump frame pairs, and then the number of abnormal jump frame pairs can be obtained as the first number.

[0067] In some embodiments, the first threshold value may be related to the shape and direction of the target road section. For a target road section with a curvature within a first value range, a corresponding first target threshold value may be set; for a target road section with a curvature within a second value range, a corresponding second target threshold value may be set. The maximum value of the first value range is less than the minimum value of the second value range, and the first target threshold value is less than the second target threshold value. In other words, for a target road section containing lane lines that change more dramatically, a larger maximum allowable threshold value for the jump amplitude may be set; whereas for a target road section containing lane lines that change more slowly, a smaller maximum allowable threshold value for the jump amplitude may be set.

[0068] In this way, corresponding target thresholds can be pre-set for different target sections. At this time, before comparing the difference value with the first threshold, the corresponding target threshold can be obtained according to the label or name of the target section as the aforementioned first threshold.

[0069] In this way, by setting different first thresholds according to different evaluation sections, the accuracy and stability of lane line recognition can be evaluated more flexibly and accurately, avoiding the possibility that the evaluation results of lane line recognition will be affected by the different shapes and directions of lane lines in different sections.

[0070] Step S250: Based on the ratio of the first number to the second number, determining an evaluation result of the lane line recognition stability of the target vehicle, wherein the lane line recognition stability is negatively correlated with the ratio.

[0071] After obtaining the first number of pairs of abnormal jump frames, we can further obtain the ratio of the first number to the total number of adjacent frames (i.e., the total number of difference values). This ratio is used to represent the proportion of abnormal jump frame pairs in all tested adjacent frames. Obviously, a larger ratio indicates a greater number of false positives in lane line recognition results, a poorer lane line recognition stability, and a lower lane line recognition accuracy.

[0072] In some embodiments, after step S250, the ratio may be further compared to a stability threshold. If the ratio is determined to be greater than the stability threshold, a prompt message is output to indicate that the target vehicle's lane recognition stability does not meet the corresponding stability requirement. Accordingly, the evaluator can optimize the lane recognition algorithm based on this prompt message to improve its accuracy and stability, thereby enhancing the safety of the target vehicle's autonomous driving function and providing more effective support for assisted driving.

[0073] In this mode, in addition to outputting prompt information, all abnormal jump frame pairs can also be output for testers to optimize the lane line recognition algorithm based on the data with false detection.

[0074] Optionally, if the ratio is less than or equal to the stability threshold, ensuring that the lane recognition stability of the target vehicle meets the corresponding stability requirement, the first threshold is lowered, and other target road sections are selected for lane recognition stability evaluation. The lane recognition algorithm is then further optimized based on the lower first threshold to further improve lane recognition accuracy and stability.

[0075] In other embodiments, after step S250, the ratios obtained by evaluating multiple other target road sections may be further obtained to obtain multiple ratios; the average of the multiple ratios is obtained as the target ratio; then, the target ratio is matched with the stability threshold. If it is determined that the target ratio is greater than the stability threshold, a prompt message is output to indicate that the lane recognition stability of the target vehicle does not meet the corresponding stability requirements. In this way, by obtaining the target ratios of the target vehicle's current lane recognition algorithm for multiple target road sections, it is possible to more comprehensively and accurately evaluate whether the lane recognition stability of the current target vehicle meets the corresponding stability requirements based on the target ratios and the stability threshold.

[0076] In this embodiment, the accuracy and stability of lane recognition results can be evaluated directly based on the difference between lane parameters identified in adjacent frames. This evaluation can be completed without relying on parameters collected by external reference sensors, reducing evaluation costs. Furthermore, after obtaining the difference between lane parameters in adjacent frames, the target vehicle's lane recognition performance is evaluated directly based on this difference, eliminating the need for post-processing or manual correction. This significantly improves lane recognition evaluation efficiency. Furthermore, the difference between multiple lane parameters can be used to achieve a more comprehensive and accurate evaluation, making the accuracy and stability of lane recognition results more reliable.

[0077] Please refer to Figure 3 , which shows a structural block diagram of a lane recognition evaluation device 300 provided in one embodiment of the present application. The device 300 may include: a recognition result acquisition module 310, a parameter acquisition module 320, a difference degree acquisition module 330, and an evaluation module 340.

[0078] The recognition result acquisition module 310 is used to obtain continuous multi-frame lane line recognition results.

[0079] The parameter acquisition module 320 is used to obtain the lane line parameters corresponding to each frame based on the lane line recognition results of each frame.

[0080] The difference degree acquisition module 330 is used to obtain the difference degree values ​​of the lane line parameters of adjacent frames according to the lane line parameters corresponding to each frame.

[0081] The evaluation module 340 is used to determine an evaluation result of the lane recognition performance of the target vehicle according to the difference degree value.

[0082] In some embodiments, the evaluation module 340 may include a quantity acquisition unit and an evaluation unit. The quantity acquisition unit may be configured to acquire the number of difference values ​​greater than a first threshold as a first quantity, and the total number of difference values ​​as a second quantity. The evaluation unit may be configured to determine an evaluation result of the lane recognition stability of the target vehicle based on a ratio of the first quantity to the second quantity, wherein the lane recognition stability is negatively correlated with the ratio.

[0083] In this manner, the lane line parameter includes multiple sub-parameters, and the difference degree value includes the difference value corresponding to each sub-parameter; the quantity acquisition unit can be specifically used to obtain the number of difference degree values ​​in which the target difference value is greater than the target parameter threshold, as the first quantity, the target difference value is the difference value corresponding to any sub-parameter, and the target parameter threshold corresponds to the sub-parameter corresponding to the target difference value.

[0084] In this manner, the parameter acquisition module 320 may be specifically configured to: acquire the absolute value of the difference between each sub-parameter of adjacent frames as the difference value corresponding to each sub-parameter.

[0085] In some embodiments, the lane line recognition evaluation device 300 may further include a prompt module. The prompt module may be configured to, after determining an evaluation result of the lane line recognition stability of the target vehicle based on the ratio of the first number to the second number, output a prompt message if the ratio is greater than a stability threshold, indicating that the lane line recognition stability of the target vehicle does not meet a corresponding stability requirement.

[0086] In some embodiments, the lane line parameters include the angle of the lane line, and the recognition result includes at least a lane line image, wherein the lane line image includes the recognized lane line. The parameter acquisition module 320 can be specifically configured to: for each lane line included in the lane line image, obtain a target portion of the lane line that is less than a first distance from the target vehicle; extract the lane line center point of the target portion to obtain multiple lane line center points; fit the multiple lane line center points to obtain a center point straight line; and obtain the angle between the center point straight line and a target straight line as the lane line angle, wherein the target straight line is the straight line in which the target vehicle is traveling when the recognition result is recognized.

[0087] In some embodiments, the recognition result acquisition module 310 can be specifically used to: obtain a road section with a curvature less than a preset curvature from a target map as the target road section; control the target vehicle to travel on the target road section, and perform lane line recognition on the target road during driving to obtain continuous multi-frame lane line recognition results.

[0088] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and modules can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0089] In several embodiments provided in this application, the coupling between modules may be electrical, mechanical or other forms of coupling.

[0090] In addition, the functional modules in the various embodiments of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules.

[0091] In summary, in the solution provided by the present application, continuous multi-frame lane line recognition results are obtained; based on the lane line recognition results of each frame, the lane line parameters corresponding to each frame are obtained; based on the lane line parameters corresponding to each frame, the difference degree value of the lane line parameters of adjacent frames is obtained; and based on the difference degree value, the evaluation result of the lane line recognition performance of the target vehicle is determined. In this way, the accuracy and stability of the lane line recognition results can be evaluated directly based on the difference degree value between the identified lane line parameters of adjacent frames. The evaluation can be completed without relying on the parameters collected by the external reference sensor, thereby reducing the evaluation cost; and after obtaining the difference degree value between the lane line parameters of adjacent frames, there is no need for post-processing or manual correction. Instead, the lane line recognition performance of the target vehicle is evaluated directly based on the difference degree value, which greatly improves the evaluation efficiency of lane line recognition.

[0092] The following will be combined Figure 4 An electronic device provided by this application is described.

[0093] Reference Figure 4 , Figure 4The following is a block diagram of the structure of an electronic device 400 provided in an embodiment of the present application. The above method provided in an embodiment of the present application can be executed by the electronic device 400. The electronic device can be an electronic terminal with a data processing function, and the electronic terminal includes but is not limited to a vehicle, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart watch, an e-book reader, etc. Of course, the electronic device can also be a server, and the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0094] The electronic device 400 in the embodiment of the present application may include one or more of the following components: a processor 401, a memory 402, and one or more applications, wherein the one or more applications may be stored in the memory 402 and configured to be executed by one or more processors 401, and the one or more programs are configured to execute the method as described in the aforementioned method embodiment.

[0095] Processor 401 may include one or more processing cores. Processor 401 utilizes various interfaces and circuits to connect various components within electronic device 400. It executes various functions and processes data for electronic device 400 by running or executing instructions, programs, code sets, or instruction sets stored in memory 402, as well as accessing data stored in memory 402. Optionally, processor 401 may be implemented using at least one of the following hardware forms: digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). Processor 401 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem handles wireless communications. It is understood that the modem may also be integrated into processor 401 and implemented separately via a communication chip.

[0096] The memory 402 may include a random access memory (RAM) or a read-only memory (ROM). The memory 402 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 402 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The data storage area may also store data created by the electronic device 400 during use (such as the various corresponding relationships described above).

[0097] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and modules can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0098] In several embodiments provided in this application, the coupling or direct coupling or communication connection between the modules shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0099] In addition, the functional modules in the various embodiments of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules.

[0100] Please refer to Figure 5 , which shows a block diagram of a computer-readable storage medium provided in an embodiment of the present application. The computer-readable medium 500 stores program code, which can be called by a processor to execute the method described in the above method embodiment.

[0101] The computer-readable storage medium 500 can be an electronic memory such as a flash memory, an EEPROM (Electrically Erasable Programmable Read-Only Memory), an EPROM, a hard disk, or a ROM. Alternatively, the computer-readable storage medium 500 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 500 has storage space for program code 510 for executing any of the method steps described above. These program codes can be read from or written to one or more computer program products. The program code 510 can be compressed, for example, in a suitable form.

[0102] In some embodiments, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the steps of each of the above method embodiments.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements 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 the present application.

Claims

1. A lane recognition evaluation method, characterized in that: The method comprises: Obtaining a road section with a curvature smaller than a preset curvature from the target map as a target road section; Controlling the target vehicle to travel on the target road section, and performing lane line recognition on the target road section during the driving process to obtain continuous multi-frame lane line recognition results; Obtain lane line parameters corresponding to each frame based on the lane line recognition results of each frame; the lane line parameters include at least one sub-parameter of the curvature of the lane line, the length of the lane line, the angle of the lane line, and the intercept between the target vehicle and the lane line; Obtaining a difference value of lane line parameters of adjacent frames according to the lane line parameters corresponding to each frame; the difference value includes a difference value corresponding to each sub-parameter; Determining a first threshold corresponding to the target road section based on the number or name of the target road section; Obtaining, as a first quantity, a number of difference degree values ​​having a target difference value greater than a target parameter threshold, wherein the target difference value is a difference value corresponding to any sub-parameter included in the lane line parameter, the target parameter threshold corresponds to the sub-parameter corresponding to the target difference value, and the total number of the difference degree values ​​is obtained as a second quantity; Based on the ratio of the first number to the second number, an evaluation result of the lane line recognition stability of the target vehicle is determined, and the lane line recognition stability is negatively correlated with the ratio.

2. The method according to claim 1, characterized in that The obtaining, based on the lane line parameters corresponding to each frame, a difference value of lane line parameters of adjacent frames, includes: The absolute value of the difference between each sub-parameter of adjacent frames is obtained as the difference value corresponding to each sub-parameter.

3. The method according to claim 1, characterized in that After determining an evaluation result of lane line recognition stability of the target vehicle based on the ratio of the first number to the second number, the method further includes: If the ratio is greater than the stability threshold, a prompt message is output, where the prompt message is used to prompt that the lane line recognition stability of the target vehicle does not meet the corresponding stability requirement.

4. The method according to any one of claims 1 to 3, characterized in that The lane line parameter includes an angle of the lane line, and the recognition result includes at least a lane line image, wherein the lane line image includes the recognized lane line; The lane line parameters corresponding to each frame are obtained according to the lane line recognition results of each frame, including: For each lane line included in the lane line image, obtaining a target portion of the lane line whose distance from the target vehicle is less than a first distance; Extracting a lane line center point of the target portion to obtain a plurality of lane line center points; Fitting the center points of the multiple lane lines to obtain a center point straight line; The angle between the center point straight line and the target straight line is obtained as the angle of the lane line, and the target straight line is the straight line of the moving direction of the target vehicle when the recognition result is recognized.

5. A lane recognition evaluation device, characterized in that: The device comprises: The recognition result acquisition module is used to obtain a road section with a curvature less than a preset curvature from the target map as a target road section; control the target vehicle to travel on the target road section, and perform lane line recognition on the target road section during driving to obtain continuous multi-frame lane line recognition results; A parameter acquisition module is used to obtain lane line parameters corresponding to each frame based on the lane line recognition results of each frame; the lane line parameters include at least one sub-parameter of the lane line curvature, lane line length, lane line angle, and the intercept between the target vehicle and the lane line; A difference degree acquisition module is used to obtain a difference degree value of the lane line parameters of adjacent frames based on the lane line parameters corresponding to each frame; the difference degree value includes a difference value corresponding to each sub-parameter; An evaluation module is used to determine a first threshold corresponding to the target road section based on the label or name of the target road section; obtain the number of difference degree values ​​where the target difference value is greater than the target parameter threshold as a first number, the target difference value is the difference value corresponding to any sub-parameter included in the lane line parameter, the target parameter threshold corresponds to the sub-parameter corresponding to the target difference value, and the total number of the difference degree values ​​is used as a second number; based on the ratio of the first number to the second number, determine the evaluation result of the lane line recognition stability of the target vehicle, and the lane line recognition stability is negatively correlated with the ratio.

6. An electronic device, characterized in that: include: one or more processors; Memory; One or more programs, wherein the 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 execute the method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores program code, which can be called by a processor to execute the method according to any one of claims 1 to 4.

Images