A method, system, device, and storage medium for detecting radar extrinsic error.

Abstract

This invention discloses a method for detecting radar extrinsic parameter errors, comprising the following steps: acquiring a radar point cloud and extracting a first point cloud containing a preset reference object from the radar point cloud; transforming the first point cloud using the extrinsic parameter to be detected to obtain a second point cloud; loading the reference point cloud and determining a mapping point in the reference point cloud corresponding to each point in the second point cloud; and detecting the accuracy of the extrinsic parameter to be detected in the reference point cloud based on the coordinates of each mapping point and several surrounding points. This invention also discloses a system, a computer device, and a readable storage medium. The proposed solution evaluates the error of radar extrinsic parameters by calculating the coordinate difference between the roadside radar point cloud and a dense point cloud map, and can test the accuracy of the calibrated radar extrinsic parameters.

Description

Technical Field

[0001] This invention relates to the field of lidar, specifically to a method, system, device, and storage medium for detecting radar extrinsic error. Background Technology

[0002] Currently, there are several methods for calibrating the extrinsic parameters of roadside lidar. For example, joint calibration can be performed using vehicle-mounted lidar and roadside lidar. By placing the vehicle-mounted lidar and roadside lidar in the same scene, ICP or NDT registration can be performed on the two pairs of point clouds obtained, and the extrinsic parameters of the roadside lidar in the map coordinate system can be obtained using the known extrinsic parameters of the vehicle-mounted lidar. Alternatively, a special calibration board can be used for calibration. However, there is currently no good method to detect whether the calibrated extrinsic parameters are accurate enough. Summary of the Invention

[0003] In view of this, in order to overcome at least one aspect of the above problems, embodiments of the present invention propose a method for detecting radar extrinsic parameter errors, comprising the following steps:

[0004] Acquire radar point cloud and extract a first point cloud of a preset reference object from the radar point cloud;

[0005] The first point cloud is transformed using the extrinsic parameters to be detected to obtain the second point cloud;

[0006] Load a reference point cloud, and determine a mapping point in the reference point cloud corresponding to each point in the second point cloud;

[0007] The accuracy of detecting the extrinsic parameters to be detected is based on the coordinates of each of the mapping points and several surrounding points in the reference point cloud.

[0008] In some embodiments, the accuracy of detecting the extrinsic parameter to be detected in the reference point cloud based on the coordinates of each of the mapped points and several surrounding points further includes:

[0009] In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the center of the mapping point is calculated with each of the mapping points as the center.

[0010] The accuracy of the extrinsic parameter to be detected is determined based on all absolute values ​​corresponding to each of the mapping points.

[0011] In some embodiments, detecting the accuracy of the extrinsic parameter to be detected based on all absolute values ​​corresponding to each mapping point further includes:

[0012] Calculate the first average of all absolute values ​​corresponding to each of the mapping points;

[0013] The second average value is calculated by summing the first average values ​​corresponding to each of the mapping points and used as the error of the external parameter to be detected;

[0014] The accuracy of the external parameter to be detected is evaluated based on the error.

[0015] In some embodiments, the first point cloud is transformed using the extrinsic parameters to be detected to obtain the second point cloud, and the method further includes:

[0016] The second point cloud is voxelized into a mesh, where the length, width, and height of the mesh are all of a first preset size.

[0017] In some embodiments, loading a reference point cloud further includes:

[0018] Load the reference point cloud based on the range of the second point cloud;

[0019] The reference point cloud is voxelized into a mesh, wherein the length, width, and height of the mesh are all of a second preset size, and the second preset size is a first preset multiple of the first preset size.

[0020] In some embodiments, the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point is calculated in the reference point cloud, with each of the mapping points as the center, further comprising:

[0021] The size of the preset range is determined according to the second preset size, wherein the preset range is a second preset multiple of the second preset size.

[0022] In some embodiments, it also includes:

[0023] The ground is used as the preset reference object.

[0024] In some embodiments, calculating the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point in the reference point cloud, with each of the mapping points as the center, further includes:

[0025] Read the Z coordinates of the mapping point and each point within the preset range, and calculate the absolute value of the difference between the coordinates based on the Z coordinates.

[0026] Based on the same inventive concept, according to another aspect of the present invention, embodiments of the present invention also provide a radar extrinsic parameter detection system, comprising:

[0027] The acquisition module is configured to acquire the radar point cloud obtained by lidar scanning and extract a first point cloud of a preset reference object from the radar point cloud.

[0028] The conversion module is configured to convert the first point cloud into a second point cloud using the extrinsic parameters to be detected.

[0029] The loading module is configured to load a reference point cloud based on the range of the second point cloud, and determine a mapping point in the reference point cloud corresponding to each point in the second point cloud;

[0030] The detection module is configured to detect the accuracy of the external parameters to be detected based on the coordinates of each of the mapping points and several surrounding points in the reference point cloud.

[0031] Based on the same inventive concept, according to another aspect of the present invention, embodiments of the present invention also provide a computer device, comprising:

[0032] At least one processor; and

[0033] The memory stores a computer program that can run on the processor, which, when executing the program, performs the steps of any of the radar extrinsic error detection methods described above.

[0034] Based on the same inventive concept, according to another aspect of the present invention, embodiments of the present invention also provide a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of any of the radar extrinsic error detection methods described above.

[0035] The present invention has one of the following beneficial technical effects: The solution proposed in this invention evaluates the error of radar extrinsic parameters by calculating the coordinate difference between the roadside radar point cloud and the dense point cloud map. This not only tests the accuracy of the calibrated radar extrinsic parameters, but also detects changes in radar pose in a timely manner. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0037] Figure 1 A schematic flowchart of a radar extrinsic parameter error detection method provided for an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of the structure of a radar extrinsic parameter detection system provided in an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of the structure of a computer device provided for an embodiment of the present invention;

[0040] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided for an embodiment of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to specific examples and the accompanying drawings.

[0042] It should be noted that all uses of "first" and "second" in the embodiments of the present invention are for the purpose of distinguishing two entities or parameters with the same name but different names. It is clear that "first" and "second" are only for the convenience of expression and should not be construed as limiting the embodiments of the present invention. Subsequent embodiments will not explain this in detail.

[0043] In an embodiment of the present invention, the purpose of extrinsic parameter calibration of the lidar is to obtain the pose of the lidar in a certain coordinate system. The pose includes translation in the x-axis, y-axis, and z-axis directions and rotation angles around the x-axis, y-axis, and z-axis in the coordinate system.

[0044] The calibration of extrinsic parameters of lidar refers to solving the relative transformation relationship between the lidar measurement coordinate system and the measurement coordinate systems of other sensors, i.e., the rotation and translation transformation matrix. The accuracy of the extrinsic parameter calibration determines the accuracy of the perception system in locating the actual pose of the lidar. In vehicle-road cooperative systems, extrinsic parameter calibration also plays a decisive role in the perception network obtaining the specific location of traffic participants in the real world.

[0045] In vehicle-road cooperative systems, LiDAR is installed at a fixed position on a roadside pole. In order to use the objects detected by the roadside LiDAR for autonomous driving, it is necessary to calibrate the extrinsic parameters of the roadside LiDAR in the map coordinate system.

[0046] According to one aspect of the present invention, embodiments of the present invention provide a method for detecting radar extrinsic parameter errors, such as... Figure 1 As shown, it may include the following steps:

[0047] S1, acquire radar point cloud and extract the first point cloud of a preset reference object from the radar point cloud;

[0048] S2, the first point cloud is transformed using the external parameters to be detected to obtain the second point cloud;

[0049] S3, Load the reference point cloud, and determine the mapping point in the reference point cloud corresponding to each point in the second point cloud;

[0050] S4, in the reference point cloud, the accuracy of the external parameter to be detected is determined based on the coordinates of each of the mapping points and several surrounding points.

[0051] The proposed solution evaluates the error of radar extrinsic parameters by calculating the coordinate difference between the roadside radar point cloud and the dense point cloud map. This not only tests the accuracy of the calibrated radar extrinsic parameters but also detects changes in radar pose in a timely manner.

[0052] In some embodiments, it also includes:

[0053] The ground is used as the preset reference object.

[0054] In some embodiments, calculating the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point in the reference point cloud, with each of the mapping points as the center, further includes:

[0055] Read the Z coordinates of the mapping point and each point within the preset range, and calculate the absolute value of the difference between the coordinates based on the Z coordinates.

[0056] Specifically, by selecting the ground as the preset reference, only the difference between the Z coordinates of the mapped point and its surrounding points can be calculated.

[0057] In some embodiments, in step S1, a radar point cloud obtained from a lidar scan is acquired, and a first point cloud of a preset reference object is extracted from the radar point cloud. Specifically, the point cloud can be segmented first, dividing it into ground point cloud and non-ground point cloud. Then, by setting a ground threshold for coordinate z, the ground point cloud and non-ground point cloud can be initially segmented, and then the ground point cloud can be filtered using the verticality of the ground normal vector.

[0058] It should be noted that the ground tilt situation has been taken into account in the ground point cloud extraction algorithm of roadside radar. As long as the scene when the image is collected is consistent with the current actual scene, it will not affect the validity of the error.

[0059] In some embodiments, the accuracy of detecting the extrinsic parameter to be detected in the reference point cloud based on the coordinates of each of the mapped points and several surrounding points further includes:

[0060] In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the center of the mapping point is calculated with each of the mapping points as the center.

[0061] The accuracy of the extrinsic parameter to be detected is determined based on all absolute values ​​corresponding to each of the mapping points.

[0062] In some embodiments, detecting the accuracy of the extrinsic parameter to be detected based on all absolute values ​​corresponding to each mapping point further includes:

[0063] Calculate the first average of all absolute values ​​corresponding to each of the mapping points;

[0064] The second average value is calculated by summing the first average values ​​corresponding to each of the mapping points and used as the error of the external parameter to be detected;

[0065] The accuracy of the external parameter to be detected is evaluated based on the error.

[0066] Specifically, such as Figure 2 As shown, traverse each point in the second point cloud. Let this point be A. Then, the point in the reference point cloud that completely overlaps with point A is A'. Using point A' as the center in the reference point cloud, obtain all points P1, P2…P' within a preset range. n Then calculate the z-coordinate of A' and P1, P2...P n The average of the absolute values ​​of the differences in the z-coordinates:

[0067]

[0068] Assuming the second point cloud grid has M points, then

[0069]

[0070] The final D value represents the error of the lidar extrinsic parameters to be determined. The smaller the error value, the more accurate the radar extrinsic parameters are.

[0071] In some embodiments, the first point cloud is transformed using the extrinsic parameters to be detected to obtain the second point cloud, and the method further includes:

[0072] The second point cloud is voxelized into a mesh, where the length, width, and height of the mesh are all of a first preset size.

[0073] Specifically, voxelizing the second point cloud reduces the number of points, thus reducing the number of points required for coordinate difference calculations. In some embodiments, the dimensions of the meshed second point cloud can be set to 0.25 meters.

[0074] In some embodiments, loading a reference point cloud further includes:

[0075] Load the reference point cloud based on the range of the second point cloud;

[0076] The reference point cloud is voxelized into a mesh, wherein the length, width, and height of the mesh are all of a second preset size, and the second preset size is a first preset multiple of the first preset size.

[0077] Specifically, the reference point cloud can be loaded from a dense point cloud map based on the range of the second point cloud, and the reference point cloud can be meshed into voxels, which can also reduce the number of points. Since the second point cloud is extracted from the radar point cloud, it is relatively sparse compared to the reference point cloud. Therefore, after meshing the reference point cloud, it is necessary not only to retain a sufficient number of points for subsequent error calculations but also to reduce the number of points. Thus, the dimensions of the reference point cloud in the mesh can be set to twice the size of the second point cloud (the larger the mesh, the smaller the points), for example, 0.5 meters.

[0078] In some embodiments, the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point is calculated in the reference point cloud, with each of the mapping points as the center, further comprising:

[0079] The size of the preset range is determined according to the second preset size, wherein the preset range is a second preset multiple of the second preset size.

[0080] Specifically, in order to optimize the amount of calculation while ensuring the accuracy of error calculation, the size of the preset range can be set to twice the size of the second preset range, for example, it can be set to 1 meter.

[0081] The proposed solution uses the ground as the true value and evaluates the error of radar extrinsic parameters by calculating the difference between the ground of the roadside radar point cloud and the ground of the dense point cloud map in the z-direction of the coordinate system. This not only tests the accuracy of the calibrated radar extrinsic parameters but also detects changes in radar pose in a timely manner.

[0082] Based on the same inventive concept, according to another aspect of the present invention, embodiments of the present invention also provide a radar extrinsic parameter detection system 400, such as... Figure 2 As shown, it includes:

[0083] The acquisition module 401 is configured to acquire radar point cloud and extract a first point cloud of a preset reference object from the radar point cloud;

[0084] The conversion module 402 is configured to convert the first point cloud into a second point cloud using the external parameters to be detected;

[0085] The loading module 403 is configured to load a reference point cloud and determine a mapping point in the reference point cloud corresponding to each point in the second point cloud;

[0086] The detection module 404 is configured to detect the accuracy of the external parameters to be detected in the reference point cloud based on the coordinates of each of the mapping points and several surrounding points.

[0087] In some embodiments, the detection module 404 is further configured to:

[0088] In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the center of the mapping point is calculated with each of the mapping points as the center.

[0089] The accuracy of the extrinsic parameter to be detected is determined based on all absolute values ​​corresponding to each of the mapping points.

[0090] In some embodiments, the detection module 405 is further configured to:

[0091] Calculate the first average of all absolute values ​​corresponding to each of the mapping points;

[0092] The second average value is calculated by summing the first average values ​​corresponding to each of the mapping points and used as the error of the external parameter to be detected;

[0093] The accuracy of the external parameter to be detected is evaluated based on the error.

[0094] In some embodiments, the conversion module 402 is further configured to:

[0095] The second point cloud is voxelized into a mesh, where the length, width, and height of the mesh are all of a first preset size.

[0096] In some embodiments, the loading module 403 is further configured to:

[0097] Load the reference point cloud based on the range of the second point cloud;

[0098] The reference point cloud is voxelized into a mesh, wherein the length, width, and height of the mesh are all of a second preset size, and the second preset size is a first preset multiple of the first preset size.

[0099] In some embodiments, the computing module 404 is further configured to:

[0100] The size of the preset range is determined according to the second preset size, wherein the preset range is a second preset multiple of the second preset size.

[0101] In some embodiments, a determining module is further included, configured to:

[0102] The ground is used as the preset reference object.

[0103] In some embodiments, the computing module 404 is further configured to:

[0104] Read the Z coordinates of the mapping point and each point within the preset range, and calculate the absolute value of the difference between the coordinates based on the Z coordinates.

[0105] The proposed solution uses the ground as the true value and evaluates the error of radar extrinsic parameters by calculating the difference between the ground of the roadside radar point cloud and the ground of the dense point cloud map in the z-direction of the coordinate system. This not only tests the accuracy of the calibrated radar extrinsic parameters but also detects changes in radar pose in a timely manner.

[0106] Based on the same inventive concept, according to another aspect of the present invention, such as Figure 3 As shown, embodiments of the present invention also provide a computer device 501, comprising:

[0107] At least one processor 520; and

[0108] The memory 510 stores a computer program 511 that can be run on a processor. When the processor 520 executes the program, it performs the steps of any of the radar extrinsic error detection methods described above.

[0109] Based on the same inventive concept, according to another aspect of the present invention, such as Figure 4 As shown, embodiments of the present invention also provide a computer-readable storage medium 601, which stores a computer program 610. When the computer program 610 is executed by a processor, it performs the steps of any of the radar extrinsic error detection methods described above.

[0110] Finally, it should be noted that those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods.

[0111] Furthermore, it should be understood that the computer-readable storage medium (e.g., memory) described herein may be volatile memory or non-volatile memory, or may include both volatile memory and non-volatile memory.

[0112] Those skilled in the art will also understand that the various exemplary logic blocks, modules, circuits, and algorithm steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the functionality of various illustrative components, blocks, modules, circuits, and steps has been generally described. Whether this functionality is implemented as software or as hardware depends on the specific application and the design constraints imposed on the system as a whole. Those skilled in the art can implement the functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the embodiments disclosed herein.

[0113] The above are exemplary embodiments disclosed in this invention. However, it should be noted that various changes and modifications can be made without departing from the scope of the embodiments of this invention as defined by the claims. The functions, steps, and / or actions of the methods according to the disclosed embodiments described herein do not need to be performed in any particular order. Furthermore, although the elements disclosed in the embodiments of this invention may be described or claimed individually, they may be understood as multiple unless explicitly limited to a singular number.

[0114] It should be understood that, as used herein, the singular form “a” is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, “and / or” refers to any and all possible combinations of one or more of the associated listed items.

[0115] The embodiment numbers disclosed in the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0116] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0117] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples. Within the framework of the invention, technical features of the above embodiments or different embodiments can be combined, and many other variations of different aspects of the invention exist, which are not provided in the details for the sake of brevity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A method for detecting radar extrinsic parameter errors, characterized in that, Includes the following steps: Acquire radar point cloud and extract a first point cloud of a preset reference object from the radar point cloud; The first point cloud is transformed using the extrinsic parameters to be detected to obtain the second point cloud; Load a reference point cloud, and determine a mapping point in the reference point cloud corresponding to each point in the second point cloud; The accuracy of detecting the extrinsic parameters to be detected is based on the coordinates of each of the mapping points and several surrounding points in the reference point cloud. The accuracy of detecting the extrinsic parameters to be detected based on the coordinates of each of the mapped points and several surrounding points in the reference point cloud further includes: In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the center of the mapping point is calculated with each of the mapping points as the center. The accuracy of the external parameter to be detected is determined based on all absolute values ​​corresponding to each mapping point, wherein the smaller the average value of all absolute values ​​corresponding to each mapping point, the more accurate the external parameter to be detected.

2. The method as described in claim 1, characterized in that, The accuracy of detecting the extrinsic parameter to be detected based on all absolute values ​​corresponding to each of the mapping points further includes: Calculate the first average of all absolute values ​​corresponding to each of the mapping points; The second average value is calculated by summing the first average values ​​corresponding to each of the mapping points and used as the error of the external parameter to be detected; The accuracy of the external parameter to be detected is evaluated based on the error.

3. The method as described in claim 1, characterized in that, The second point cloud is obtained by transforming the first point cloud using the extrinsic parameters to be detected, and further includes: The second point cloud is voxelized into a mesh, where the length, width, and height of the mesh are all of a first preset size.

4. The method as described in claim 3, characterized in that, Loading a reference point cloud, further including: Load the reference point cloud based on the range of the second point cloud; The reference point cloud is voxelized into a mesh, wherein the length, width, and height of the mesh are all of a second preset size, and the second preset size is a first preset multiple of the first preset size.

5. The method as described in claim 4, characterized in that, In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point is calculated, with each mapping point as the center. This further includes: The size of the preset range is determined according to the second preset size, wherein the preset range is a second preset multiple of the second preset size.

6. The method as described in claim 1, characterized in that, Also includes: The ground is used as the preset reference object.

7. The method as described in claim 6, characterized in that, In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the central mapping point is calculated, with each mapping point as the center, further including: Read the Z coordinates of the mapping point and each point within the preset range, and calculate the absolute value of the difference between the coordinates based on the Z coordinates.

8. A radar extrinsic parameter detection system, characterized in that, include: The acquisition module is configured to acquire the radar point cloud obtained by lidar scanning and extract a first point cloud of a preset reference object from the radar point cloud. The conversion module is configured to convert the first point cloud into a second point cloud using the extrinsic parameters to be detected. The loading module is configured to load a reference point cloud based on the range of the second point cloud, and determine a mapping point in the reference point cloud corresponding to each point in the second point cloud; The detection module is configured to detect the accuracy of the external parameters to be detected in the reference point cloud based on the coordinates of each of the mapping points and several surrounding points; The detection module is also configured as follows: In the reference point cloud, the absolute value of the difference between the coordinates of each point within a preset range and the center of the mapping point is calculated with each of the mapping points as the center. The accuracy of the external parameter to be detected is determined based on all absolute values ​​corresponding to each mapping point, wherein the smaller the average value of all absolute values ​​corresponding to each mapping point, the more accurate the external parameter to be detected.

9. A computer device, comprising: At least one processor; as well as A memory storing a computer program executable on the processor, characterized in that the processor executes the program by performing the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it performs the steps of the method as described in any one of claims 1-7.

Images