Parking space detection precision test method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of vehicle performance testing, and provides a parking space detection precision testing method and device, electronic equipment and a storage medium. The method comprises the steps of determining current target pose information of a to-be-tested vehicle when the to-be-tested vehicle reaches a preset final position and the variation of pose information is smaller than a preset value in a preset time period, and obtaining test data of a to-be-tested parking space based on a laser radar and a camera on the to-be-tested vehicle; and according to the pre-stored actual parking space data of the to-be-tested parking space and the test data and target pose information of the to-be-tested parking space corresponding to the preset test times, determining a detection precision evaluation result of the to-be-tested vehicle for the to-be-tested parking space, thereby solving the problems of complex operation and tedious steps of testing the detection precision of the parking space in the prior art. And the parking space detection precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle performance testing, and in particular to a method, device, electronic equipment and storage medium for testing parking space detection accuracy. Background Art

[0002] With the continuous development of autonomous driving technology and intelligent transportation systems, how to accurately and quickly detect parking spaces has become a key issue. The research on parking space detection can not only effectively improve the utilization rate of parking spaces, but also alleviate the problem of limited parking space resources.

[0003] Traditional parking space detection methods mainly rely on ground marking lines, such as horizontal and vertical parking spaces. However, this method has certain limitations. For example, in low light or blurred marking lines, the detection accuracy will be significantly affected.

[0004] In order to ensure the accuracy and reliability of parking space detection before the vehicle is released, it is necessary to conceive a comprehensive real-vehicle testing method for parking space detection accuracy. Summary of the invention

[0005] In view of this, the embodiments of the present application provide a method, device, electronic device and storage medium for testing parking space detection accuracy to solve the problem that the testing operation of parking space detection accuracy in the prior art is complicated and the steps are cumbersome.

[0006] According to a first aspect of an embodiment of the present application, a method for testing parking space detection accuracy is provided, comprising:

[0007] In the process of the vehicle to be tested driving from a preset starting position to a preset end position in the parking space test scene, the position information of the vehicle to be tested is obtained in real time; the parking spaces to be tested in the parking space test scene include horizontal parking spaces and vertical parking spaces;

[0008] When the vehicle to be tested reaches a preset end position and the change in the posture information within a preset period is less than a preset value, the current target posture information of the vehicle to be tested is determined, and the test data of the parking space to be tested is obtained based on the laser radar and camera on the vehicle to be tested;

[0009] The detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested is determined according to the pre-stored actual parking space data of the parking space to be tested and the test data and target posture information of the parking space to be tested corresponding to the preset test times.

[0010] According to a second aspect of an embodiment of the present application, a parking space detection accuracy testing device is provided, comprising:

[0011] The acquisition module is configured to acquire the position information of the vehicle to be tested in real time during the process of the vehicle to be tested driving from a preset starting position to a preset end position in the parking space test scene; the parking spaces to be tested in the parking space test scene include horizontal parking spaces and vertical parking spaces;

[0012] A determination module is configured to determine the current target posture information of the vehicle to be tested when the vehicle to be tested reaches a preset end position and the change in posture information within a preset time period is less than a preset value, and obtain test data of the parking space to be tested based on the laser radar and camera on the vehicle to be tested;

[0013] The evaluation module is configured to determine the detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested based on the pre-stored actual parking space data of the parking space to be tested and the test data and target posture information of the parking space to be tested corresponding to the preset test times.

[0014] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0015] According to a fourth aspect of an embodiment of the present application, a readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0016] Compared with the prior art, the embodiments of the present application have the following beneficial effects: by obtaining the vehicle's posture information in real time during the process of the vehicle to be tested traveling from the preset starting position to the preset end position, the posture information is used to monitor the motion state of the vehicle to be tested during the entire test process. After the vehicle to be tested reaches the preset end position, and the change in the posture information within the preset time period is less than the preset value, it can be considered that the vehicle has stabilized at this time, and its final posture information can be recorded as the target posture information. By setting a threshold value of the change within a preset time period, it can be ensured that when recording the target posture information, the vehicle to be tested has stopped and the position is stable, which helps to improve the accuracy of data acquisition. When the vehicle to be tested stops and records the target posture information, the test data of the parking space to be tested obtained based on the laser radar and the camera is obtained, and the test data of the parking space to be tested is the direct basis for evaluating the accuracy of parking space detection. Based on the actual parking space data of the parking space to be tested stored in advance and the test data and target posture information obtained through multiple tests, the accuracy of parking space detection is quantified by comparing the difference between the test data and the actual parking space data, and the accuracy of the parking space detection of the vehicle to be tested is evaluated. The above-mentioned test method is easy to operate and the scene setting is simple. It can specifically quantify the indicators of the vehicle's parking space detection performance, and can very intuitively obtain the level of spatial parking space detection of the vehicle to be tested. It solves the problem of complex test operation and cumbersome steps for parking space detection accuracy in the prior art, and improves the parking space detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 It is a flow chart of a method for testing parking space detection accuracy provided in an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of a horizontal parking space test scenario in an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of a vertical parking space test scenario in an embodiment of the present application;

[0021] Figure 4 It is a schematic diagram of displaying the current target posture information of the vehicle to be tested and the test data of the parking space to be tested in an embodiment of the present application;

[0022] Figure 5 is a schematic diagram of the results of a horizontal parking space test scenario in an embodiment of the present application;

[0023] Figure 6 is a schematic diagram of a parking space detection accuracy evaluation result in an embodiment of the present application;

[0024] Figure 7 is a schematic diagram of another horizontal parking space test scenario in an embodiment of the present application;

[0025] Figure 8 It is a flow chart of another method for testing parking space detection accuracy provided in an embodiment of the present application;

[0026] Fig. 9 is another schematic diagram of a horizontal parking space test scenario in an embodiment of the present application;

[0027] Fig.10 is another schematic diagram of a vertical parking space test scenario in an embodiment of the present application;

[0028] Fig.11 It is a structural schematic diagram of a parking space detection accuracy testing device provided in an embodiment of the present application;

[0029] Fig.12 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0031] Figure 1 It is a flow chart of a method for testing parking space detection accuracy provided in an embodiment of the present application.

[0032] like Figure 1 As shown, the test method for the parking space detection accuracy includes:

[0033] Step 101, in the process of the vehicle to be tested driving from a preset starting position to a preset end position in a parking space test scene, the position information of the vehicle to be tested is obtained in real time; the parking spaces to be tested in the parking space test scene include horizontal parking spaces and vertical parking spaces.

[0034] Specifically, in order to test the maturity and stability of parking detection performance, the accuracy test of parking detection should be added before the release of the vehicle. In order to ensure a standardized test environment, repeatable testing, and controllable test conditions, it is necessary to create a test scenario, and build horizontal and vertical parking test scenarios to comprehensively test the detection performance of different types of parking spaces.

[0035] Specific parking space test scenarios may include: horizontal parking space test scenarios and vertical parking space test scenarios. As an example, Figure 2 This is one of the schematic diagrams of the horizontal parking space test scenario, refer to Figure 2 In the horizontal parking space test scenario, the width of the parking space to be tested (the distance between the parking space line corresponding to point C and the parking space line corresponding to point D) can be set to the average length of the vehicle to be tested plus 80 cm. The vehicle to be tested starts from the preset starting position and stops at the preset end position. The first obstacle vehicle and the second obstacle vehicle are parked on both sides of the parking space to be tested. Figure 3 This is a schematic diagram of the vertical parking test scenario, see Figure 3 In the vertical parking space test scenario, the width of the parking space to be tested (the distance between the parking space line corresponding to point C and the parking space line corresponding to point D) can be set to the average width of the vehicle to be tested plus 80 cm. The vehicle to be tested starts from the preset starting position and stops at the preset end position. The first obstacle vehicle and the second obstacle vehicle are parked on both sides of the parking space to be tested.

[0036] In some embodiments, the whole vehicle is powered on, the vehicle system is activated to ensure that all electronic systems (such as sensors, controllers, etc.) of the vehicle under test are in working condition, and an external computer can be connected to the intelligent driving controller of the vehicle under test through a dedicated interface (such as CAN bus, USB or other proprietary interface), allowing engineers to access the control system of the vehicle under test through a computer during the test, view real-time data and debug the system, and enter the engineering development mode. In the engineering development mode, engineers are allowed to access detailed information and debugging tools inside the vehicle under test. In this mode, it is easier to obtain vehicle status information for debugging and testing.

[0037] The vehicle under test turns on the automatic parking assist system (APA) to search for parking spaces. The vehicle drives at idle speed (vehicle speed <15km / h) from the preset starting position to the preset end position. By driving at a lower speed to simulate the actual driving scenario, the vehicle under test can be given more time to accurately detect parking spaces and reduce false detections or missed detections caused by high-speed driving.

[0038] In the process of the vehicle under test driving from the preset starting position to the preset end position in the parking test scene, the posture information of the vehicle under test is obtained in real time, which can be used to monitor the movement state of the vehicle under test during the entire test process, record the position and direction of the vehicle under test at different time points, ensure that the vehicle drives along the preset path, and calculate the change in the posture of the vehicle under test over time, so as to evaluate the driving stability of the vehicle under test.

[0039] The position and posture information of the vehicle to be tested includes the coordinates of the vehicle to be tested and the direction angle θ of the vehicle to be tested. The coordinate system where the coordinates of the vehicle to be tested are located takes the preset starting position as the origin.

[0040] Including different types of parking spaces (horizontal and vertical) in the test scenarios can more comprehensively evaluate the vehicle's detection capabilities in different parking scenarios, helping to ensure that the vehicle can cope with various complex parking environments in actual use.

[0041] Step 102, when the vehicle to be tested reaches a preset terminal position and the change in posture information within a preset time period is less than a preset value, determine the current target posture information of the vehicle to be tested, and obtain test data of the parking space to be tested based on the laser radar and camera on the vehicle to be tested.

[0042] In some embodiments, when the vehicle to be tested reaches a preset end position and the change in posture information within a preset time period is less than a preset value, on the one hand, it ensures that each test ends at the same position, thereby ensuring the consistency of the test conditions, and can provide stable recording points to ensure the stability of data acquisition and avoid errors caused by changes in vehicle position; on the other hand, by setting a preset time period (such as 30 seconds) and requiring that the change in the vehicle's posture information during this period is less than a preset value (such as 0.1 meters or 0.1 degrees), the vehicle to be tested can be stationary when recording data, avoiding data fluctuations caused by slight movements of the vehicle to be tested, thereby improving the reliability of data acquisition.

[0043] Record the posture information of the vehicle under test in a stable state, that is, the current target posture information (x / y / θ) of the vehicle under test. The target posture information can be used as benchmark data for subsequent data analysis and comparison.

[0044] At the same time, the laser radar and camera on the vehicle to be tested are used to obtain the test data of the parking space to be tested. The test data of the parking space to be tested includes the depth, width and test coordinates of the preset points of the parking space to be tested. For example, the current target posture information of the vehicle to be tested and the test data of the parking space to be tested can be as follows: Figure 4As shown, at this time, the current target posture information Pose of the vehicle to be tested is (17.89, -0.83, -0.0833), the test width I of the parking space to be tested is 5.1, the test depth d of the parking space to be tested is 2.17, and the test coordinates of the preset point P0 of the parking space to be tested are (15.36, -2.28).

[0045] Among them, when the parking space test scene is a horizontal parking space test scene, the obtained test data and target posture information are data under the horizontal space parking space test scene; when the parking space test scene is a vertical parking space test scene, the obtained test data and target posture information are data under the vertical parking space test scene.

[0046] In this way, by obtaining the target posture information and the test data of the parking space to be tested, all necessary information for evaluating the detection performance of the vehicle to be tested can be obtained, providing a basis for subsequent accuracy evaluation.

[0047] Step 103 , determining a detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested based on the pre-stored actual parking space data of the parking space to be tested and the test data and target posture information of the parking space to be tested corresponding to the preset test times.

[0048] Specifically, the pre-stored actual parking space data of the parking space to be tested includes the actual depth, actual width, and actual distance between the preset point and the preset end position of the parking space to be tested. The actual parking space data can be used as a benchmark for comparison with the test data.

[0049] Through a preset number of parking space detection tests, multiple test data and target posture information of the parking space to be tested are obtained, more data is accumulated, the reliability and representativeness of the data are improved, and statistical analysis is helpful. The preset number of tests can be 20 times, 10 times on each side.

[0050] By comparing the differences between the test data and the actual parking space data, for example, calculating error statistics (such as pass rate, standard deviation, etc.), quantifying the detection performance, obtaining the detection accuracy evaluation result of the vehicle under test for the parking space under test, and evaluating the accuracy of parking space detection.

[0051] The above test method is easy to operate and the scene is simple to build. It can specifically quantify the indicators of parking space detection performance, and can easily obtain the level of spatial parking space detection based on the fusion solution of the laser radar and camera on the vehicle to be tested, and quickly determine whether it meets the indicator requirements. In addition, the above test method requires little manpower investment, the test tools are common and easy to obtain, and the test method is low in cost.

[0052] In some embodiments, the test data of the parking space to be tested includes the test width, test depth and test coordinates of the preset point of the parking space to be tested; the actual parking space data includes the actual width, actual depth and actual distance between the preset point and the preset end position of the parking space to be tested;

[0053] According to the pre-stored actual parking space data of the parking space to be tested and the test data and target posture information of the parking space to be tested corresponding to the preset test times, the detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested is determined, including:

[0054] For each test in the preset number of tests, a width deviation is obtained according to the test width and the actual width;

[0055] According to the test depth and the actual depth, the depth deviation is obtained;

[0056] Determine the test distance between the preset end point and the corresponding preset point according to the target posture information corresponding to the preset end point and the test coordinates of the preset point corresponding to the preset end point, and obtain the distance deviation according to the test distance and the actual distance;

[0057] The detection accuracy evaluation result is obtained based on the preset width deviation, depth deviation and distance deviation corresponding to the predicted number of times.

[0058] Specifically, the test data of the parking space to be tested includes the test width, test depth and test coordinates of the preset point of the parking space to be tested. Figure 5 , Figure 5 This is a schematic diagram of the results of the horizontal parking space test scenario, where Figure 5 Points P1, P2, P3 and P4 are the four corner points of the ultrasonic horizontal parking space obtained by measurement, Obj1 is the first obstacle, Obj2 is the second obstacle, point P1 is the preset point obtained by measurement, I represents the test width of the parking space to be measured, d represents the test depth of the parking space to be measured, and the test coordinates of the preset point P1 obtained by measurement are pose(x / y / θ), L P1 represents the measured distance between the preset end position and the preset point P1, which is measured when the rear axle center of the vehicle to be tested coincides with the preset end position. P1′ represents the actual preset point of the parking space to be tested, I′ represents the actual width of the parking space to be tested, d′ represents the actual depth of the parking space to be tested, and L P1′ Indicates the actual distance between the preset point and the preset end position.

[0059] The test data is obtained by the vehicle under test through its sensors during the test process, reflecting the vehicle under test's ability to perceive the parking space. The actual data is the real state of the parking space under test, which is used to compare with the test data to evaluate the accuracy of detection. In each test, the difference between the test width I and the actual width I' is calculated to obtain the width deviation ΔI, which is used to evaluate the accuracy of width detection; the difference between the test depth d and the actual depth d' is calculated to obtain the depth deviation Δd, which is used to evaluate the accuracy of depth detection; by calculating the measured distance L between the preset end position and the corresponding preset point P1 The actual distance L P1v The difference between them is used to obtain the distance deviation, which is used to evaluate the accuracy of position detection; it can also indirectly reflect the vehicle's ability to recognize the parking space boundary and internal structure.

[0060] The actual distance between the preset point and the preset end position is a preset value, which is determined when the test scene is constructed. Specifically, the actual distance L between the preset point and the preset end position is P1′ It can be 10 meters; the test distance L between the preset point of the parking space to be tested and the target posture information corresponding to the preset end position P1 It can be calculated by the following formula: L P1 =(pose(x)-P1(x))*cosθ+(pose(y)-P1(y))*sinθ;

[0061] Where pose(x) is the x-axis value corresponding to the preset end point position, pose(y) is the y-axis value corresponding to the preset end point position, P1(y) is the y-axis value corresponding to the P1 point position, P1(x) is the x-axis value corresponding to the P1 point position, and θ is the direction angle corresponding to the preset end point position.

[0062] By summarizing the width deviation, depth deviation and distance deviation of each test, a comprehensive detection accuracy evaluation result is obtained. The accuracy of the test on the vehicle to be tested is detected by determining several key parameters of the parking space to be tested, which improves the convenience and accuracy of the precision test.

[0063] In some embodiments, the detection accuracy evaluation result is obtained according to the preset width deviation, depth deviation and distance deviation corresponding to the prediction times, including:

[0064] Determine the width qualification rate according to the preset width deviation range and the preset individual width deviations;

[0065] Calculate the standard deviation of the preset width deviations to obtain the corresponding width standard deviation;

[0066] According to the preset depth deviation range and the preset individual depth deviations, the depth qualification rate is obtained;

[0067] Calculate the standard deviation of the preset depth deviations to obtain the corresponding depth standard deviation;

[0068] According to the preset distance deviation range and the preset individual distance deviation, the distance qualification rate is obtained;

[0069] Calculate the standard deviation of the preset distance deviations to obtain the corresponding distance standard deviation;

[0070] The width pass rate, width standard deviation, depth pass rate, depth standard deviation, distance pass rate and distance standard deviation are determined as the detection accuracy evaluation results.

[0071] Specifically, by comparing the width deviation of each test with a preset width deviation range, it is determined whether the width detection is within the allowable error range, and the proportion of the width deviation falling within the preset range in the preset number of tests is calculated to obtain the width qualification rate.

[0072] By calculating the standard deviation of the width deviation in multiple tests, the corresponding width standard deviation is obtained to evaluate the fluctuation range of the width deviation. The smaller the standard deviation, the more stable the width detection result.

[0073] By comparing the depth deviation of each test with the preset depth deviation range, it is determined whether the depth detection is within the allowable error range, and the proportion of depth deviations falling within the preset range in the preset number of tests is calculated to obtain the depth qualification rate.

[0074] By calculating the standard deviation of the depth deviation in multiple tests, the corresponding depth standard deviation is obtained to evaluate the fluctuation range of the depth deviation. The smaller the standard deviation, the more stable the depth detection result.

[0075] By comparing the distance deviation of each test with the preset distance deviation range, it is determined whether the distance detection is within the allowable error range, and the proportion of distance deviations falling within the preset range in the preset number of tests is calculated to obtain the distance qualification rate.

[0076] By calculating the standard deviation of distance deviation in multiple tests, the corresponding distance standard deviation is obtained to evaluate the fluctuation range of distance deviation. The smaller the standard deviation, the more stable the distance detection result.

[0077] The parking space detection accuracy evaluation results can be Figure 6 As shown, through Figure 6 The statistical graphs shown show the width deviation and distance deviation, in Figure 6In the parking space deviation statistics chart, horizontal parking_R can represent the width deviation of the horizontal parking space to be tested from the right side of the parking space to be tested, horizontal parking_L can represent the width deviation of the horizontal parking space to be tested from the left side of the parking space to be tested, vertical parking_R can represent the width deviation of the vertical parking space to be tested from the right side of the parking space to be tested, and vertical parking_L can represent the width deviation of the vertical parking space to be tested from the left side of the parking space to be tested. The horizontal axis represents the number of tests. For example, when the horizontal axis coordinate is 1, it means that the data is the data of the first test. 20 cm on the vertical axis is the upper limit of the width deviation of the parking space, and -15 cm on the vertical axis is the lower limit of the width deviation of the parking space. For example, when the width deviation of the parking space is greater than +20, or when the width deviation of the parking space is less than -15, it can be considered that the width deviation is not within the preset width deviation range.

[0078] exist Figure 6 In the p1 (preset point) deviation statistics chart, horizontal parking_R can represent the distance deviation when detecting the horizontal parking space to be tested from the right side of the parking space to be tested, horizontal parking_L can represent the distance deviation when detecting the horizontal parking space to be tested from the left side of the parking space to be tested, vertical parking_R can represent the distance deviation when detecting the vertical parking space to be tested from the right side of the parking space to be tested, and vertical parking_L can represent the distance deviation when detecting the vertical parking space to be tested from the left side of the parking space to be tested. The horizontal axis represents the number of tests. For example, when the horizontal axis is 2, it means that the data is the data of the second test. The 15 cm on the vertical axis is the upper limit of the distance deviation of the parking space. For example, when the distance deviation is greater than +15, it can be considered that the distance deviation is not within the preset distance deviation range.

[0079] In addition, the parking space detection accuracy evaluation results are shown in the following table.

[0080] Table 1: Parking space detection accuracy evaluation results

[0081]

[0082] Among them, the horizontal parking space R is the evaluation result of detecting the horizontal parking space to be tested from the right side of the parking space to be tested, the horizontal parking space L is the evaluation result of detecting the horizontal parking space to be tested from the left side of the parking space to be tested, the vertical parking space R is the evaluation result of detecting the vertical parking space to be tested from the right side of the parking space to be tested, and the vertical parking space L is the evaluation result of detecting the vertical parking space to be tested from the left side of the parking space to be tested.

[0083] In some embodiments, a first position and a second position are preset in the parking space test scene, wherein when the preset starting position is the first position, the preset end position is the second position; when the preset starting position is the second position, the preset end position is the first position; the preset point corresponds to the preset end position, and the actual distances between the first position and the second position and the corresponding preset points are the same;

[0084] The test data includes test data when the preset end position is the first position and test data when the preset end position is the second position;

[0085] The target posture information includes posture information when the preset end position is a first position and posture information when the preset end position is a second position.

[0086] Specifically, by presetting the first position and the second position in the parking space test scenario, the test can be performed in two directions to ensure the comprehensiveness and bidirectionality of the test. Through the bidirectional test, the performance of the system in different driving directions can be better evaluated.

[0087] refer to Figure 7 , when the first position is the preset starting position, the preset end position is the second position, and the preset point corresponding to the preset end position is point D. The coordinate system at this time takes the first position as the origin, and the test data at this time is the test data when the preset end position is the second position, including the test width, test depth and test coordinates of the preset point of the parking space to be tested when the vehicle to be tested reaches the second position. The target posture information at this time is the target posture information of the vehicle to be tested when the vehicle to be tested reaches the second position. When the second position is the preset starting position, the preset end position is the first position, and the preset point corresponding to the preset end position is point C. The coordinate system at this time takes the second position as the origin, and the test data at this time is the test data when the preset end position is the first position, including the test width, test depth and test coordinates of the preset point of the parking space to be tested when the vehicle to be tested reaches the first position. The target posture information at this time is the target posture information of the vehicle to be tested when the vehicle to be tested reaches the first position.

[0088] In addition, in some embodiments, the parking space detection accuracy test process can be as follows: Figure 8 As shown, before the test begins, the vehicle state to be tested, the test site and the test tools need to be prepared.

[0089] Among them, preparation of the state of the vehicle to be tested is a key step before the test, which requires ensuring that the vehicle to be tested is in the best working condition, which may include whether sensors such as radar and camera are working normally; test site preparation involves the coordination and preparation of the test site, including determining the number and type of parking spaces in the test scenario (such as horizontal parking spaces and vertical parking spaces). At the same time, it is also necessary to ensure that other facilities or objects in the test site will not interfere with the test, site confirmation (ensuring that the test site meets the test requirements), and test vehicle parking position confirmation (the vehicle is parked at the preset starting position); test tool preparation may include confirming the version information of the intelligent driving software carried by the vehicle to be tested, and whether the software has the required functions.

[0090] At the beginning of the test, for the horizontal parking space test, the vehicle to be tested starts from the first position and searches for an empty parking space on the left. The vehicle to be tested tries to find an empty horizontal parking space. When the vehicle to be tested reaches the second position, it stops and records the conditional output data at this time. These data can reflect the detection results of the parking space by the vehicle to be tested. The above process is repeated many times (such as 10 tests on each side) to collect enough data for subsequent analysis. Optionally, the average value of the final result can be taken. For the vertical parking space test, at the beginning of the test, the vehicle to be tested starts from the first position and searches for an empty parking space on the left. The vehicle to be tested tries to find an empty vertical parking space. When the vehicle to be tested reaches the second position, it stops and records the conditional output data at this time. These data can reflect the detection results of the parking space by the vehicle to be tested. The above process is repeated many times (such as 10 tests on each side) to collect enough data for subsequent analysis. Optionally, the average value of the final result can be taken.

[0091] In some embodiments, test data of the parking space to be tested is obtained based on the laser radar and camera on the vehicle to be tested, including: obtaining parking space point cloud data detected by the laser radar; obtaining parking space image data captured by the camera; fusing the parking space point cloud data and the parking space image data to obtain parking space fusion feature data; based on the parking space fusion feature data, performing parking space detection through a parking space detection model to obtain test data of the parking space to be tested.

[0092] Specifically, lidar can generate a three-dimensional point cloud map of the environment. The parking space point cloud data detected by lidar can provide very accurate distance measurement, and the parking space image data captured by the camera can provide rich color and texture information.

[0093] Before fusing the parking space point cloud data and the parking space image data, the parking space point cloud data can be preprocessed, including filtering to remove noise points, segmenting the point cloud data, extracting the ground and objects parallel to the ground, etc.; in addition, the parking space image data can also be preprocessed, including image segmentation, feature extraction, etc., identifying lane lines, parking space markings, etc.

[0094] The pre-processed parking spot cloud data and parking spot image data are unified into the same coordinate system. Through the feature matching algorithm, the corresponding feature points in the parking spot cloud data and parking spot image data are found. The complete environment model is constructed by combining the three-dimensional modeling capability of the LiDAR and the visual information of the camera. The distance and position information provided by the LiDAR is fused with the visual information provided by the camera to determine the position, size and shape of the parking space and obtain the parking space fusion feature data. The parking space fusion feature data is input into the trained parking space detection model. The fusion feature data is automatically processed by the model to realize the rapid detection of the parking space and obtain the test data of the parking space to be tested.

[0095] The combined use of lidar and cameras can make full use of their respective advantages and improve the accuracy of parking space detection.

[0096] In some embodiments, before obtaining the posture information of the vehicle to be tested in real time, it also includes: controlling the vehicle to be tested to be stationary at a preset starting position, and controlling the vehicle to be tested to power off the entire vehicle to initialize the posture information of the vehicle to be tested.

[0097] Specifically, by keeping the vehicle under test stationary at a preset starting position, it can be ensured that the position and posture of the vehicle are known and consistent at the beginning of each test, which helps to reduce the error introduced by the different initial positions of the vehicle and improve the accuracy of subsequent position and posture information acquisition. Powering off the entire vehicle can clear temporary data and status information that may exist in the system of the vehicle under test (such as the trajectory information of the vehicle under test), avoiding temporary data from interfering with the initialization of posture information. At the same time, the power-off operation can be regarded as a restart of the vehicle system, which helps to restore the initial state of the system and ensure that all sensors and control systems are restarted in a clean state at the beginning of the test.

[0098] In some embodiments, the parking space test scene includes a first obstacle vehicle, a second obstacle vehicle, two longitudinal marking lines, and a transverse marking line;

[0099] The horizontal marking line is the first parking space line corresponding to the vehicle entry side of the parking space to be tested;

[0100] The two longitudinal marking lines are respectively two second parking space lines on the parking space to be measured and perpendicular to the first parking space line;

[0101] The first obstacle vehicle and the second obstacle vehicle are respectively located outside the two longitudinal marking lines of the parking space to be measured, and when the parking space to be measured is a horizontal parking space, the front or rear of the first obstacle vehicle and the front or rear of the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines; when the parking space to be measured is a vertical parking space, the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the fronts of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines;

[0102] The vehicle to be tested and the first obstacle vehicle are respectively located on both sides of the transverse marking line, and the driving track of the vehicle to be tested is at a preset distance from the transverse marking line.

[0103] Specifically, Fig. 9It is a schematic diagram of the horizontal parking space test scenario. The horizontal marking line is parallel to the straight line from the starting position to the end position. Optionally, the horizontal marking line is the first parking space line corresponding to the vehicle entry side of the parking space to be tested; the vertical marking line is a line perpendicular to the driving trajectory of the vehicle to be tested. Optionally, the two vertical marking lines are two second parking space lines on the parking space to be tested that are perpendicular to the first parking space line, which can be used to define the depth of the parking space.

[0104] The first obstacle vehicle is set on the left side of the horizontal parking space to be tested, that is, on the left side of the longitudinal marking line, and is located on both sides of the transverse marking line with the vehicle to be tested. The length side of the body of the first obstacle vehicle coincides with the transverse marking line, and the front or rear of the vehicle coincides with the longitudinal marking line, indicating that the first obstacle vehicle defines the left boundary of the horizontal parking space to be tested. The second obstacle vehicle is set on the right side of the horizontal parking space to be tested, that is, on the right side of another longitudinal marking line, and is located on both sides of the transverse marking line with the vehicle to be tested. The length side of the body of the second obstacle vehicle coincides with the transverse marking line, and the front or rear of the vehicle coincides with the longitudinal marking line, indicating that the second obstacle vehicle defines the right boundary of the horizontal parking space to be tested.

[0105] Maintaining a preset distance between the driving trajectory of the vehicle to be tested and the horizontal marking line can simulate the distance between the vehicle and the parking space boundary during actual parking and ensure the safety of the test. Specifically, the preset distance can be 1 meter.

[0106] For example, as an example, Fig.10 This is a schematic diagram of the vertical parking space test scenario. The setting principles of the horizontal marking lines and the vertical marking lines are the same as the horizontal parking space lines, which will not be repeated here.

[0107] The first obstacle vehicle is set on the left side of the vertical parking space, that is, on the left side of the longitudinal marking line, and is located on both sides of the transverse marking line with the vehicle to be tested. The front of the first obstacle vehicle coincides with the transverse marking line, and the length of the vehicle body coincides with the longitudinal marking line, indicating that the first obstacle vehicle defines the left boundary of the vertical parking space. The second obstacle vehicle is set on the right side of the vertical parking space, that is, on the right side of another longitudinal marking line, and is located on both sides of the transverse marking line with the vehicle to be tested. The front of the second obstacle vehicle coincides with the transverse marking line, and the length of the vehicle body coincides with the longitudinal marking line, indicating that the second obstacle vehicle defines the right boundary of the vertical parking space.

[0108] By setting fixed obstacle car positions and marking lines, the conditions for each test can be consistent, making the test results comparable. The standardized test scenario allows researchers to repeat the same test in order to verify the effectiveness of improvement measures.

[0109] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

[0110] The following are device embodiments of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0111] Fig.11 Schematic diagram of a parking space detection accuracy testing device provided in an embodiment of the present application. Fig.11 As shown, the test device for the parking space detection accuracy includes:

[0112] The acquisition module 1101 is configured to acquire the position information of the vehicle to be tested in real time during the process of the vehicle to be tested driving from a preset starting position to a preset end position in the parking space test scene; the parking space to be tested in the parking space test scene includes a horizontal parking space and a vertical parking space;

[0113] The determination module 1102 is configured to determine the current target posture information of the vehicle to be tested, and obtain the test data of the parking space to be tested based on the laser radar and the camera on the vehicle to be tested, when the vehicle to be tested reaches the preset end position and the change amount of the posture information within the preset time period is less than the preset value;

[0114] The evaluation module 1103 is configured to determine the detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested based on the pre-stored actual parking space data of the parking space to be tested and the test data and target posture information of the parking space to be tested corresponding to the preset test times.

[0115] According to the technical solution provided in the embodiment of the present application, based on the actual parking space data of the pre-stored parking space to be tested and the test data and target posture information obtained through multiple tests, by comparing the difference between the test data and the actual parking space data, it is possible to quickly determine whether the parking space detection performance meets the index requirements, and the above-mentioned test method steps are easy to operate, the scene construction is simple, the indicators of the vehicle's parking space detection performance can be specifically quantified, and the level of detection of the spatial parking space by the vehicle to be tested can be obtained very intuitively, thereby solving the problem of complex test operation and cumbersome steps for the parking space detection accuracy in the prior art, and improving the parking space detection accuracy.

[0116] In some embodiments, the test data of the parking space to be tested includes the test width, test depth and test coordinates of the preset point of the parking space to be tested; the actual parking space data includes the actual width, actual depth and actual distance between the preset point and the preset end position of the parking space to be tested; the evaluation module 1103 is configured to obtain the width deviation according to the test width and the actual width for each test in the preset number of tests; obtain the depth deviation according to the test depth and the actual depth; determine the test distance between the preset end position and the corresponding preset point according to the target posture information corresponding to the preset end position and the test coordinates of the preset point corresponding to the preset end position, and obtain the distance deviation according to the test distance and the actual distance; obtain the detection accuracy evaluation result according to the preset width deviation, depth deviation and distance deviation corresponding to the predicted number of times.

[0117] In some embodiments, the evaluation module 1103 is configured to determine the width qualification rate based on a preset width deviation range and preset width deviations; perform standard deviation calculation on the preset width deviations to obtain the corresponding width standard deviation; obtain the depth qualification rate based on a preset depth deviation range and preset depth deviations; perform standard deviation calculation on the preset depth deviations to obtain the corresponding depth standard deviation; obtain the distance qualification rate based on a preset distance deviation range and preset distance deviations; perform standard deviation calculation on the preset distance deviations to obtain the corresponding distance standard deviation; determine the width qualification rate, width standard deviation, depth qualification rate, depth standard deviation, distance qualification rate and distance standard deviation as the detection accuracy evaluation result.

[0118] In some embodiments, a first position and a second position are pre-set in the parking space test scene, wherein when the preset starting position is the first position, the preset end position is the second position; when the preset starting position is the second position, the preset end position is the first position; the preset point corresponds to the preset end position, and the first position and the second position are respectively the same as the actual distances between the corresponding preset points; the test data includes test data when the preset end position is the first position and test data when the preset end position is the second position; the target posture information includes posture information when the preset end position is the first position and posture information when the preset end position is the second position.

[0119] In some embodiments, the determination module 1102 is configured to obtain parking space point cloud data detected by the laser radar; obtain parking space image data captured by the camera; fuse the parking space point cloud data and the parking space image data to obtain parking space fusion feature data; perform parking space detection through a parking space detection model based on the parking space fusion feature data to obtain test data of the parking space to be tested.

[0120] In some embodiments, the acquisition module 1101 is configured to control the vehicle to be tested to be stationary at a preset starting position and control the vehicle to be tested to power off to initialize the posture information of the vehicle to be tested before acquiring the posture information of the vehicle to be tested in real time.

[0121] In some embodiments, the parking space test scene includes a first obstacle vehicle, a second obstacle vehicle, two longitudinal marking lines, and a transverse marking line;

[0122] The horizontal marking line is the first parking space line corresponding to the vehicle entry side of the parking space to be tested;

[0123] The two longitudinal marking lines are respectively two second parking space lines on the parking space to be measured and perpendicular to the first parking space line;

[0124] The first obstacle vehicle and the second obstacle vehicle are respectively located outside the two longitudinal marking lines of the parking space to be measured, and when the parking space to be measured is a horizontal parking space, the front or rear of the first obstacle vehicle and the front or rear of the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines; when the parking space to be measured is a vertical parking space, the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the fronts of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines;

[0125] The vehicle to be tested and the first obstacle vehicle are respectively located on both sides of the transverse marking line, and the driving track of the vehicle to be tested is at a preset distance from the transverse marking line.

[0126] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0127] Fig.12 Schematic diagram of an electronic device 12 provided in an embodiment of the present application. Fig.12 As shown, the electronic device 12 of this embodiment includes: a processor 1201, a memory 1202, and a computer program 1203 stored in the memory 1202 and executable on the processor 1201. When the processor 1201 executes the computer program 1203, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 1201 executes the computer program 1203, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0128] The electronic device 12 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 12 may include, but is not limited to, a processor 1201 and a memory 1202. Those skilled in the art will appreciate that Fig.12 The electronic device 12 is merely an example and does not limit the electronic device 12 , and may include more or fewer components than those shown in the figure, or different components.

[0129] Processor 1201 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0130] The memory 1202 may be an internal storage unit of the electronic device 12, for example, a hard disk or memory of the electronic device 12. The memory 1202 may also be an external storage device of the electronic device 12, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card (FlashCard), etc. equipped on the electronic device 12. The memory 1202 may also include both an internal storage unit of the electronic device 12 and an external storage device. The memory 1202 is used to store computer programs and other programs and data required by the electronic device.

[0131] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0132] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. Computer-readable storage media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0133] 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. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for testing parking space detection accuracy, characterized in that: include: Acquire the position information of the vehicle under test in real time while the vehicle under test is driving from a preset starting position to a preset end position in the parking space test scene; The parking spaces to be tested in the parking space test scenario include horizontal parking spaces and vertical parking spaces; When the vehicle to be tested reaches the preset end position and the change amount of the posture information within a preset time period is less than a preset value, determine the current target posture information of the vehicle to be tested, and obtain test data of the parking space to be tested based on the laser radar and camera on the vehicle to be tested; The detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested is determined according to the pre-stored actual parking space data of the parking space to be tested, the test data of the parking space to be tested corresponding to the preset test times, and the target posture information.

2. The method according to claim 1, characterized in that The test data of the parking space to be tested includes the test width and test depth of the parking space to be tested and the test coordinates of the preset point of the parking space to be tested; the actual parking space data includes the actual width and actual depth of the parking space to be tested and the actual distance between the preset point and the preset end point; The step of determining the detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested based on the pre-stored actual parking space data of the parking space to be tested, the test data of the parking space to be tested corresponding to the preset test times, and the target posture information comprises: For each test in the preset number of tests, obtaining a width deviation according to the test width and the actual width; Obtaining a depth deviation according to the test depth and the actual depth; Determine a test distance between the preset end position and the corresponding preset point according to the target posture information corresponding to the preset end position and the test coordinates of the preset point corresponding to the preset end position, and obtain a distance deviation according to the test distance and the actual distance; The detection accuracy evaluation result is obtained according to the preset width deviation, depth deviation and distance deviation corresponding to the number of predictions.

3. The method according to claim 2, characterized in that The detecting accuracy evaluation result is obtained according to the preset width deviation, depth deviation and distance deviation corresponding to the predicted number of times, including: Determining a width qualification rate according to a preset width deviation range and a preset number of said width deviations; Calculating the standard deviation of the preset width deviations to obtain the corresponding width standard deviation; Obtaining a depth qualification rate according to a preset depth deviation range and a preset number of depth deviations; Calculating the standard deviation of the preset depth deviations to obtain the corresponding depth standard deviation; Obtaining a distance qualification rate according to a preset distance deviation range and a preset number of said distance deviations; Calculating the standard deviation of the preset distance deviations to obtain the corresponding distance standard deviation; The width qualified rate, the width standard deviation, the depth qualified rate, the depth standard deviation, the distance qualified rate and the distance standard deviation are determined as the detection accuracy evaluation result.

4. The method according to any one of claims 1 to 3, characterized in that A first position and a second position are pre-set in the parking space test scene, wherein when the preset starting position is the first position, the preset end position is the second position; when the preset starting position is the second position, the preset end position is the first position; the preset point of the parking space to be tested corresponds to the preset end position, and the actual distances between the first position and the second position and the corresponding preset points are the same; The test data includes test data when the preset end position is a first position and test data when the preset end position is a second position; The target posture information includes posture information when the preset end position is a first position and posture information when the preset end position is a second position.

5. The method according to claim 1, characterized in that The step of obtaining the test data of the parking space to be tested based on the laser radar and camera on the vehicle to be tested includes: Obtaining parking spot cloud data detected by the laser radar; Acquire parking space image data captured by the camera; Fusing the parking space point cloud data and the parking space image data to obtain parking space fusion feature data; According to the parking space fusion feature data, parking space detection is performed through a parking space detection model to obtain test data of the parking space to be tested.

6. The method according to claim 1, characterized in that Before obtaining the position and posture information of the vehicle to be tested in real time, the method further includes: The vehicle to be tested is controlled to be stationary at the preset starting position, and the vehicle to be tested is controlled to be powered off to initialize the position and posture information of the vehicle to be tested.

7. The method according to claim 1, characterized in that The parking space test scene includes a first obstacle vehicle, a second obstacle vehicle, two longitudinal marking lines and a transverse marking line; The transverse marking line is a first parking space line corresponding to the vehicle entry side of the parking space to be measured; The two longitudinal marking lines are respectively two second parking space lines on the parking space to be measured and perpendicular to the first parking space line; The first obstacle vehicle and the second obstacle vehicle are respectively located outside the two longitudinal marking lines of the parking space to be measured, and when the parking space to be measured is a horizontal parking space, the front or rear of the first obstacle vehicle and the front or rear of the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines; when the parking space to be measured is a vertical parking space, the length sides of the bodies of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the corresponding longitudinal marking lines, and the fronts of the first obstacle vehicle and the second obstacle vehicle respectively coincide with the transverse marking lines and are located on the same side of the transverse marking lines; The vehicle to be tested and the first obstacle vehicle are respectively located on both sides of the transverse marking line, and the driving track of the vehicle to be tested is at a preset distance from the transverse marking line.

8. A parking space detection accuracy testing device, characterized in that: include: The acquisition module is configured to acquire the posture information of the vehicle to be tested in real time during the process of the vehicle to be tested driving from a preset starting position to a preset end position in the parking space test scene; the parking spaces to be tested in the parking space test scene include horizontal parking spaces and vertical parking spaces; A determination module is configured to determine the current target posture information of the vehicle to be tested, and obtain the test data of the parking space to be tested based on the laser radar and the camera on the vehicle to be tested, when the vehicle to be tested reaches the preset end position and the change amount of the posture information within a preset time period is less than a preset value; The evaluation module is configured to determine the detection accuracy evaluation result of the vehicle to be tested for the parking space to be tested based on the pre-stored actual parking space data of the parking space to be tested, the test data of the parking space to be tested corresponding to the preset test times, and the target posture information.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images