A method, device and related equipment for obtaining multi-radar data calibration parameters

Abstract

The present application discloses a method, apparatus and related equipment for obtaining calibration parameters of multiple radar data. The method obtains point cloud data collected by multiple radars on the same three non-parallel surfaces. According to the installation position of the radar, the point cloud data collected by each radar is corrected to obtain the corrected point cloud data corresponding to each radar. For each radar, the point cloud data corresponding to each of the three non-parallel surfaces is determined from the corresponding corrected point cloud data. The point cloud data of each surface is fitted to obtain the corresponding calibration plane, and the calibration parameters corresponding to each radar are determined using the obtained calibration plane. The present application uses each radar to collect data on the same three non-parallel surfaces, and uses the collected point cloud data of the three surfaces to fit the calibration planes corresponding to the three surfaces in each radar. The calibration planes corresponding to the same surface in different radars are used to determine the calibration parameters corresponding to each radar, so as to facilitate the calibration of multiple radar data.

Description

Technical Field

[0001] The present application relates to the field of surveying and mapping technology, and in particular to a method, device and related equipment for obtaining calibration parameters of multiple radar data. Background Art

[0002] With the rapid development of science and technology, various fields have begun to use LiDAR to collect data. For example, in the field of autonomous driving, LiDAR is one of the important sensors that can collect environmental information around autonomous vehicles in real time. When using LiDAR for data collection, multiple LiDARs with relatively fixed positions are generally used simultaneously. However, each LiDAR has its own coordinate system, so the position data measured by each LiDAR is also generated based on its own coordinate system. This makes the position data measured by each LiDAR possibly inconsistent, making it impossible to combine the position data measured by each LiDAR. Therefore, how to obtain data calibration parameters for multiple LiDARs in order to calibrate the data of multiple LiDARs has always been a problem that people are concerned about. Summary of the Invention

[0003] In view of this, the present application provides a method, device, equipment and readable storage medium for obtaining multi-radar data calibration parameters, so as to facilitate the acquisition of multi-radar data calibration parameters, thereby realizing the calibration of multi-radar data.

[0004] In order to achieve the above objectives, the following solutions are proposed:

[0005] A method for obtaining multi-radar data calibration parameters, comprising:

[0006] Obtain point cloud data collected by multiple radars on the same three non-parallel surfaces;

[0007] According to the installation position of the radar, the point cloud data collected by each radar is corrected to obtain the corrected point cloud data corresponding to each radar;

[0008] For each radar, determine the point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data, and use the point cloud data of each surface to fit the corresponding calibration plane;

[0009] The obtained calibration plane is used to determine the calibration parameters corresponding to each radar.

[0010] Optionally, the point cloud data of each surface is used to fit the corresponding calibration plane, including:

[0011] Using the point cloud data of each surface, the corresponding fitting plane is obtained by fitting;

[0012] For each fitting plane, points whose distance to the fitting plane is less than a preset threshold are selected from the corresponding point cloud data to form a candidate point set, and the candidate point set is used to fit the calibration plane corresponding to each fitting plane.

[0013] Optionally, before obtaining the calibration parameters corresponding to each radar using the obtained calibration plane, the method further includes:

[0014] For each calibration plane, select points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold to form a new candidate point set, and determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold;

[0015] If yes, then use the new candidate point set to fit a new calibration plane, and return to the step of selecting points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold;

[0016] If not, the calibration plane corresponding to each fitting plane is obtained.

[0017] Optionally, the obtained calibration plane is used to determine calibration parameters corresponding to each radar, including:

[0018] Determine one radar from among all radars as a reference radar;

[0019] For each of the three non-parallel planes, select a calibration plane corresponding to a reference radar from the corresponding calibration planes as a reference plane, and determine an attitude adjustment amount in the process of adjusting the calibration plane corresponding to each radar to coincide with the reference plane, as a calibration sub-parameter corresponding to each radar;

[0020] For each radar, the three calibration sub-parameters corresponding to the radar are used to form the calibration parameters corresponding to each radar.

[0021] Optionally, the point cloud data collected by each radar is corrected according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar, including:

[0022] From all radars, one radar is determined as a reference radar;

[0023] For each radar, the relative position relationship with the reference radar is determined according to the installation position of the radar, and the point cloud data collected by the radar is corrected according to the relative position relationship to obtain the corrected point cloud data corresponding to each radar.

[0024] Optionally, also include:

[0025] For the point cloud data collected by each radar, the point cloud data collected by each radar is corrected according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar;

[0026] For the corrected point cloud data of each radar, the calibration parameters corresponding to the radar are used to calibrate the corrected point cloud data.

[0027] A multi-radar data calibration parameter acquisition device, comprising:

[0028] A point cloud data acquisition unit, configured to acquire point cloud data collected by multiple radars on the same three non-parallel surfaces;

[0029] The point cloud data correction unit is used to correct the point cloud data collected by each radar according to the installation position of the radar, and obtain the corrected point cloud data corresponding to each radar;

[0030] a calibration plane determination unit, configured to determine, for each radar, point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data, and to fit the corresponding calibration plane using the point cloud data of each surface;

[0031] The calibration parameter determination unit is used to determine the calibration parameters corresponding to each radar using the obtained calibration plane.

[0032] Optionally, also include:

[0033] A correction unit is used to correct the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar;

[0034] The calibration unit is used to calibrate the corrected point cloud data of each radar using the calibration parameters corresponding to the radar.

[0035] A multi-radar data calibration parameter acquisition device, comprising: a memory and a processor;

[0036] The memory is used to store programs;

[0037] The processor is used to execute the program to implement the various steps of the aforementioned method for obtaining multi-radar data calibration parameters.

[0038] A readable storage medium stores a computer program thereon, which, when executed by a processor, implements the various steps of the aforementioned method for acquiring multiple radar data calibration parameters.

[0039] As can be seen from the above technical solutions, the embodiments of the present application provide a method, device, equipment and readable storage medium for obtaining calibration parameters of multiple radar data, which obtains point cloud data collected by multiple radars on the same three non-parallel surfaces, and corrects the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar. For each radar, the point cloud data corresponding to each of the three non-parallel surfaces is determined from the corresponding corrected point cloud data, and the point cloud data of each surface is used to fit the corresponding calibration plane, and the calibration parameters corresponding to each radar are determined using the obtained calibration plane. The present application uses each radar to collect data on the same three non-parallel surfaces, and uses the collected point cloud data of the three surfaces to fit the calibration planes corresponding to the three surfaces in each radar, and uses the calibration planes corresponding to the same surface in different radars to finally determine the calibration parameters corresponding to each radar, so as to facilitate the calibration of data from multiple radars. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0041] Figure 1 A flow chart of a method for obtaining multi-radar data calibration parameters provided in an embodiment of the present application;

[0042] Figure 2 A schematic diagram of the structure of a multi-radar data calibration parameter acquisition device provided in an embodiment of the present application;

[0043] Figure 3 This is a hardware structure block diagram of a multi-radar data calibration parameter acquisition device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0045] Figure 1 A flowchart of a method for obtaining multi-radar data calibration parameters provided in an embodiment of the present application, the method may include the following steps:

[0046] Step S100: Obtain point cloud data collected by multiple radars on the same three non-parallel surfaces.

[0047] Specifically, multiple radars that need to be calibrated are used to collect data on the same three non-parallel surfaces, obtaining point cloud data for the three surfaces collected by each radar that needs to be calibrated. The three surfaces are non-parallel and contain orthogonal information. For example, two intersecting walls and the ground can also be used as the three surfaces for point cloud data collection in this application.

[0048] Step S101: Correct the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar.

[0049] Specifically, a full-view long-range radar can be installed on the roof of an autonomous vehicle. However, since the full-view long-range radar is installed on the roof, a blind spot will appear near the vehicle body. Therefore, short-range blind spot radars can be installed around the vehicle to collect data at locations that the full-view long-range radar cannot observe.

[0050] Therefore, in order to meet the needs of data collection, the installation position of each radar may be different. When the point cloud data collected by each radar are processed uniformly, there will be a large position deviation. Therefore, the point cloud data collected by each radar can be corrected according to the installation position of the radar, so that the same areas between the point cloud data collected by each radar overlap as much as possible, reducing the position deviation and facilitating subsequent use.

[0051] Step S102: For each radar, determine the point cloud data corresponding to three non-parallel surfaces from the corresponding corrected point cloud data, and use the point cloud data of each surface to fit the corresponding calibration plane.

[0052] Specifically, after obtaining the corrected point cloud data corresponding to each radar, the point cloud data corresponding to each of the three non-parallel surfaces can be determined from the corrected point cloud data corresponding to each radar. The point cloud data corresponding to each of the three non-parallel surfaces can be determined by first determining a projection plane orthogonal to each surface, projecting the point cloud data onto the projection plane. Since the point cloud data is a discrete point, each point on the projection plane can be scaled up, and then the points corresponding to the surface orthogonal to the projection plane can be determined from the scaled-up two-dimensional image. Using the point cloud data corresponding to each surface, the corresponding calibration plane can be fitted. The least squares method can be used to fit the calibration plane.

[0053] Step S103: Using the obtained calibration plane, determine the calibration parameters corresponding to each radar.

[0054] Specifically, in the above steps, each radar can obtain its own three corresponding calibration planes, where the three calibration planes correspond to the three non-parallel planes. Using the obtained calibration planes, by adjusting the calibration planes, the calibration parameters corresponding to each radar can be finally determined.

[0055] In the above embodiment, a method for obtaining calibration parameters of multiple radar data is provided, which obtains point cloud data collected by multiple radars on the same three non-parallel surfaces, corrects the point cloud data collected by each radar according to the installation position of the radar, and obtains the corrected point cloud data corresponding to each radar. For each radar, the point cloud data corresponding to each of the three non-parallel surfaces is determined from the corresponding corrected point cloud data, and the point cloud data of each surface is used to fit the corresponding calibration plane. The calibration parameters corresponding to each radar are determined using the obtained calibration plane. This application uses each radar to collect data on the same three non-parallel surfaces, and uses the point cloud data of the three surfaces collected to fit the calibration planes corresponding to the three surfaces in each radar. The calibration planes corresponding to the same surface in different radars are used to finally determine the calibration parameters corresponding to each radar, so as to facilitate the calibration of data from multiple radars.

[0056] In some embodiments of the present application, the process of correcting the point cloud data collected by each radar according to the installation position of the radar in step S101 to obtain the corrected point cloud data corresponding to each radar may include:

[0057] S11. From all radars, determine one radar as a reference radar.

[0058] Specifically, a radar can be randomly selected from all radars as a reference radar, or a radar with the largest acquisition range can be selected as a reference radar. The specific selection of which radar is used as the reference radar does not affect the implementation of the various embodiments of the present application.

[0059] S12. For each radar, determine its relative positional relationship with the reference radar based on its installation position, and correct the point cloud data collected by the radar based on the relative positional relationship to obtain corrected point cloud data corresponding to each radar.

[0060] Specifically, after determining the reference radar, the relative positional relationship between each radar and the reference radar can be determined based on the radar's installation location. The point cloud data collected by the radar can then be corrected based on this relative positional relationship. This relative positional relationship can include relative distance, relative height, and other factors. During the correction process, the point cloud data can be rotated to offset the entire point cloud data.

[0061] In the above embodiment, since the point cloud data collected by each radar is collected according to its own coordinate system, when the point cloud data of each radar are brought together, the point cloud data between the radars may be misaligned. Therefore, the relative position relationship between each radar and the reference radar can be used to correct the point cloud data of each radar, so that the point cloud data of each radar can be spliced ​​together for subsequent use.

[0062] During the actual data collection process, noise may appear in the collected data due to signal fluctuations or harsh environments. When fitting a plane using point cloud data with noise, if the noise deviates far from the true plane, the fitted plane may be significantly affected, resulting in a large deviation between the fitted plane and the true plane. Based on this, in some embodiments of the present application, an optional fitting method may be provided, which may include:

[0063] S21. Using the point cloud data of each surface, a corresponding fitting plane is obtained by fitting.

[0064] Specifically, the above embodiment can obtain point cloud data for each surface. Since the point cloud data is composed of a number of discrete points, the corresponding fitting plane can be obtained by fitting the discrete points in the point cloud data. The process of fitting the plane using the discrete points in the point cloud data can be performed using the least squares method.

[0065] S22. For each fitting plane, select points from the corresponding point cloud data whose distance to the fitting plane is less than a preset threshold to form a candidate point set, and use the candidate point set to fit a calibration plane corresponding to each fitting plane.

[0066] Specifically, in the above steps, three fitting planes are obtained from the point cloud data of each radar. For each fitting plane, points whose distance from the fitting plane is less than a preset threshold are selected from the corresponding point cloud data. These selected points are then used to form a candidate point set. In this case, all points in the candidate point set are less than the preset threshold from the fitting plane, thus removing noise points with large deviations. Therefore, the calibration plane obtained by fitting the candidate point set is closer to the actual situation than the previous fitted plane.

[0067] In the above embodiment, after fitting the corresponding fitting plane using the point cloud data of each surface, points whose distance to the fitting plane is less than a preset threshold can be screened out for each fitting plane, noise points collected by the radar can be removed, and a quadratic fitting can be performed to obtain a calibration plane corresponding to each fitting plane. This reduces the interference of noise points on the plane fitting to a certain extent, making the obtained calibration plane closer to the real plane.

[0068] Since point cloud data may contain noise, when fitting a calibration plane using point cloud data, the noise may have a significant impact on the calibration plane fitting, causing the calibration plane to differ significantly from the actual plane, affecting data calibration. Based on this, some embodiments of the present application provide a method for iteratively updating the calibration plane. Before obtaining the calibration parameters corresponding to each radar using the obtained calibration plane in step S103, the method may further include:

[0069] S31. For each calibration plane, select points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold to form a new candidate point set, and determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than the preset threshold.

[0070] Specifically, three calibration planes can be obtained from the point cloud data of each radar. For each calibration plane, points whose distance from the calibration plane is less than a preset threshold can be selected from the corresponding candidate point set. These selected points are then used to form a new candidate point set. After obtaining the new candidate point set, the number of points in the new candidate point set can be compared with the number of points in the corresponding candidate point set to determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than the preset threshold. If so, S32 is executed; if not, S33 is executed.

[0071] For example, a calibration plane corresponds to a candidate point set, and the candidate point set contains 100 discrete points. Among them, there are 80 points in the candidate point set corresponding to the calibration plane whose distance to the calibration plane is less than a preset threshold, and when the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is greater than or equal to 90%, the iterative update is stopped. Therefore, points whose distance to the calibration plane is less than the preset threshold are selected from the corresponding candidate point set to form a new candidate point set containing 80 discrete points. Since the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than the preset threshold, that is, 80 / 100=80%<90%, S32 can be executed; if there are 91 points in the candidate point set corresponding to the calibration plane whose distance to the calibration plane is less than the preset threshold, then 90 / 100=91%>90%, so S33 can be executed.

[0072] S32. Using the new candidate point set, a new calibration plane is obtained by fitting.

[0073] Specifically, since in the above steps, it can be determined that the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than the preset threshold, the new candidate point set can be used to fit a new calibration plane, and return to execute S31, and select points from the corresponding candidate point set whose distance to the calibration plane is less than the preset threshold.

[0074] S33. Obtain a calibration plane corresponding to each fitting plane.

[0075] Specifically, when it is determined that the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is not less than a preset threshold, a calibration plane corresponding to each fitting plane can be obtained.

[0076] In the above embodiment, by continuously fitting with a new candidate point set to obtain a new calibration plane, the noise points collected by the radar can be continuously removed, so that the calibration plane obtained by fitting is closer to the real plane, thereby ensuring the accuracy of data calibration to a certain extent.

[0077] Furthermore, the present application can not only use the above-mentioned method to remove noise in the point cloud data collected by the radar, but also use the following optional methods to remove noise.

[0078] The first one is the median filtering method.

[0079] Specifically, for each radar, the point cloud data corresponding to three non-parallel surfaces are determined, and the statistical median of each point in the point cloud data of each surface is determined, so that the statistical median is used to filter and remove noise points in the point cloud data.

[0080] The second method is cluster elimination method.

[0081] Specifically, for each radar, the point cloud data corresponding to three non-parallel surfaces are determined, and each point in the point cloud data of each surface is determined. The average distance from each point to its k nearest points is calculated. The distances of all points in the point cloud can form a Gaussian distribution. By giving a given mean and variance, noise can be removed.

[0082] By using the above-mentioned denoising method, the noise points collected by the radar can be removed, so that when the plane is fitted using point cloud data, the calibration plane obtained by fitting is closer to the real plane, thereby ensuring the accuracy of data calibration to a certain extent.

[0083] In some embodiments of the present application, the process of determining the calibration parameters corresponding to each radar using the obtained calibration plane in step S103 is introduced. This process may include:

[0084] S41. Determine a radar from all radars as a reference radar.

[0085] Specifically, a radar can be randomly selected from all radars as a reference radar, or the radar with the largest acquisition area can be selected as the reference radar. The specific selection of which radar is used as the reference radar does not affect the implementation of the various embodiments of the present application.

[0086] S42. For each of the three non-parallel surfaces, select the calibration plane corresponding to the reference radar from the corresponding calibration plane as the reference plane, and determine the attitude adjustment amount in the process of adjusting the calibration plane corresponding to each radar to coincide with the reference plane, as the calibration sub-parameter corresponding to each radar.

[0087] Specifically, the point cloud data for each radar includes calibration planes corresponding to three non-parallel surfaces. For each of the three non-parallel surfaces, the calibration plane corresponding to the reference radar can be selected from the corresponding calibration planes as the reference plane. After determining the reference plane, the attitude adjustment amount required to adjust the calibration plane corresponding to each radar to coincide with the reference plane is determined as the calibration sub-parameter corresponding to each radar. The attitude adjustment amount can include a rotation angle and an offset.

[0088] For example, assume there are two radars, one of which is used as a reference radar and the other as a calibration radar. The three non-parallel surfaces are surface A, surface B, and surface C. For surface A, the calibration plane corresponding to surface A is selected from the two radars, and the calibration plane corresponding to surface A in the reference radar is used as the reference plane. The calibration plane corresponding to surface A in the calibration radar is adjusted so that the calibration plane corresponding to surface A in the calibration radar coincides with the calibration plane corresponding to surface A in the reference radar. During the adjustment process, the attitude adjustment amount generated may include a rotation angle and an offset. The same method can be used for surfaces B and C, and finally three sets of attitude adjustment amounts are obtained.

[0089] S43. For each radar, use the three calibration sub-parameters corresponding to the radar to form the calibration parameters corresponding to each radar.

[0090] Specifically, through the above steps, three calibration sub-parameters corresponding to each radar can be obtained. Among them, the three calibration sub-parameters can include the rotation angle and the translation. Using the three calibration sub-parameters corresponding to each radar, the calibration parameters corresponding to each radar can be composed.

[0091] In the above embodiment, since the three surfaces are not parallel to each other, the calibration sub-parameters of each dimension of the point cloud data collected by each radar in the three-dimensional space can be obtained by adjusting the calibration planes corresponding to the three surfaces to coincide with the corresponding reference planes. The calibration parameters composed of the calibration sub-parameters of the three dimensions can be used to calibrate the point cloud data from three dimensions, thereby improving the accuracy of the calibration.

[0092] After determining the calibration parameters corresponding to each radar, the point cloud data collected by each radar can be corrected according to the installation positions of the radars, and then calibrated using the calibration parameters corresponding to each radar, so that the point cloud data measured by each radar can be combined and used.

[0093] Specifically, after determining the calibration parameters corresponding to each radar, the point cloud data collected by each radar can be corrected based on the radar's installation position to obtain the corrected point cloud data corresponding to each radar. After obtaining the corrected point cloud data corresponding to each radar, the corrected point cloud data can be calibrated using the calibration parameters corresponding to the radar.

[0094] In the above embodiment, after the point cloud data is corrected according to the installation positions between the radars, the corrected point cloud data can be calibrated using the calibration parameters corresponding to each radar, so that the point cloud data measured by each radar can be combined and used.

[0095] The following describes a multi-radar data calibration parameter acquisition device provided in an embodiment of the present application. The multi-radar data calibration parameter acquisition device described below and the multi-radar data calibration parameter acquisition method described above can refer to each other.

[0096] Figure 2 A schematic diagram of the structure of a multi-radar data calibration parameter acquisition device provided in an embodiment of the present application, the multi-radar data calibration parameter acquisition device may include:

[0097] The point cloud data acquisition unit 10 is used to acquire point cloud data collected by multiple radars on the same three non-parallel surfaces;

[0098] The point cloud data correction unit 20 is used to correct the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar;

[0099] The calibration plane determination unit 30 is used to determine the point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data for each radar, and to fit the corresponding calibration plane using the point cloud data of each surface;

[0100] The calibration parameter determination unit 40 is used to determine the calibration parameters corresponding to each radar using the obtained calibration plane.

[0101] In the above embodiment, a multi-radar data calibration parameter acquisition device is provided, wherein a point cloud data acquisition unit 10 acquires point cloud data collected by multiple radars on the same three non-parallel surfaces, a point cloud data correction unit 20 corrects the point cloud data collected by each radar according to the installation position of the radar, and obtains the corrected point cloud data corresponding to each radar, a calibration plane determination unit 30 determines the point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data for each radar, and uses the point cloud data of each surface to fit a corresponding calibration plane, and a calibration parameter determination unit 40 uses the obtained calibration plane to determine the calibration parameters corresponding to each radar. This application uses each radar to collect data on the same three non-parallel surfaces, and uses the collected point cloud data of the three surfaces to fit the calibration planes corresponding to the three surfaces in each radar, and uses the calibration planes corresponding to the same surface in different radars to finally determine the calibration parameters corresponding to each radar, so as to facilitate the calibration of data from multiple radars.

[0102] Optionally, the multi-radar data calibration parameter acquisition device may further include:

[0103] A correction unit is used to correct the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar;

[0104] The calibration unit is used to calibrate the corrected point cloud data of each radar using the calibration parameters corresponding to the radar.

[0105] Optionally, the calibration plane determining unit 30 performs a process of fitting the point cloud data of each surface to obtain a corresponding calibration plane, which may include:

[0106] Using the point cloud data of each surface, the corresponding fitting plane is obtained by fitting;

[0107] For each fitting plane, points whose distance to the fitting plane is less than a preset threshold are selected from the corresponding point cloud data to form a candidate point set, and the candidate point set is used to fit the calibration plane corresponding to each fitting plane.

[0108] Optionally, the multi-radar data calibration parameter acquisition device may further include:

[0109] A candidate point set updating unit is used to select, for each calibration plane, points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold to form a new candidate point set, and to determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold;

[0110] an iterative fitting unit, configured to, when a ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold, fit a new calibration plane using the new candidate point set, and return to the step of selecting points from the corresponding candidate point set whose distance to the calibration plane is less than the preset threshold;

[0111] The calibration plane obtaining unit is configured to obtain a calibration plane corresponding to each fitting plane if the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is not less than a preset threshold.

[0112] Optionally, the calibration parameter determination unit 40 performs a process of determining the calibration parameters corresponding to each radar using the obtained calibration plane, which may include:

[0113] Determine one radar from among all radars as a reference radar;

[0114] For each of the three non-parallel planes, select a calibration plane corresponding to a reference radar from the corresponding calibration planes as a reference plane, and determine an attitude adjustment amount in the process of adjusting the calibration plane corresponding to each radar to coincide with the reference plane, as a calibration sub-parameter corresponding to each radar;

[0115] For each radar, the three calibration sub-parameters corresponding to the radar are used to form the calibration parameters corresponding to each radar.

[0116] Optionally, the point cloud data correction unit 20 corrects the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar, which may include:

[0117] From all radars, one radar is determined as a reference radar;

[0118] For each radar, the relative position relationship with the reference radar is determined according to the installation position of the radar, and the point cloud data collected by the radar is corrected according to the relative position relationship to obtain the corrected point cloud data corresponding to each radar.

[0119] The present application also provides a multi-radar data calibration parameter acquisition device. Figure 3 The hardware structure diagram of the multi-radar data calibration parameter acquisition device is shown. Figure 3 ,The hardware structure of the multi-radar data calibration parameter acquisition device may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4;

[0120] In the embodiment of the present application, the number of the processor 1, the communication interface 2, the memory 3, and the communication bus 4 is at least one, and the processor 1, the communication interface 2, and the memory 3 communicate with each other through the communication bus 4;

[0121] The processor 1 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention;

[0122] The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory;

[0123] The memory stores a program, and the processor can call the program stored in the memory, and the program is used to: implement each processing flow in the aforementioned multi-radar data calibration parameter acquisition method.

[0124] An embodiment of the present application further provides a storage medium, which may store a program suitable for execution by a processor, wherein the program is used to implement each processing flow in the aforementioned multi-radar data calibration parameter acquisition method.

[0125] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0126] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The various embodiments can be combined with each other, and the same or similar parts can be referenced to each other.

[0127] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for obtaining multi-radar data calibration parameters, characterized in that: include: Obtain point cloud data collected by multiple radars on the same three non-parallel surfaces; According to the installation position of the radar, the point cloud data collected by each radar is corrected to obtain the corrected point cloud data corresponding to each radar; For each radar, determine the point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data, and use the point cloud data of each surface to fit the corresponding calibration plane; Using the obtained calibration plane, determine the calibration parameters corresponding to each radar; Before using the obtained calibration plane to obtain the calibration parameters corresponding to each radar, the following steps are also included: For each calibration plane, select points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold to form a new candidate point set, and determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold; If yes, then use the new candidate point set to fit a new calibration plane, and return to the step of selecting points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold; If not, the calibration plane corresponding to each fitting plane is obtained.

2. The method according to claim 1, characterized in that Using the point cloud data of each surface, the corresponding calibration plane is fitted, including: Using the point cloud data of each surface, the corresponding fitting plane is obtained by fitting; For each fitting plane, points whose distance to the fitting plane is less than a preset threshold are selected from the corresponding point cloud data to form a candidate point set, and the candidate point set is used to fit the calibration plane corresponding to each fitting plane.

3. The method according to claim 1, characterized in that Using the obtained calibration plane, the calibration parameters corresponding to each radar are determined, including: Determine one radar from among all radars as a reference radar; For each of the three non-parallel planes, select a calibration plane corresponding to a reference radar from the corresponding calibration planes as a reference plane, and determine an attitude adjustment amount in the process of adjusting the calibration plane corresponding to each radar to coincide with the reference plane, as a calibration sub-parameter corresponding to each radar; For each radar, the three calibration sub-parameters corresponding to the radar are used to form the calibration parameters corresponding to each radar.

4. The method according to claim 1, wherein According to the installation position of the radar, the point cloud data collected by each radar is corrected to obtain the corrected point cloud data corresponding to each radar, including: From all radars, one radar is determined as a reference radar; For each radar, the relative position relationship with the reference radar is determined according to the installation position of the radar, and the point cloud data collected by the radar is corrected according to the relative position relationship to obtain the corrected point cloud data corresponding to each radar.

5. The method according to any one of claims 1 to 4, characterized in that Also includes: For the point cloud data collected by each radar, the point cloud data collected by each radar is corrected according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar; For the corrected point cloud data of each radar, the calibration parameters corresponding to the radar are used to calibrate the corrected point cloud data.

6. A device for acquiring multi-radar data calibration parameters, characterized in that: include: A point cloud data acquisition unit, configured to acquire point cloud data collected by multiple radars on the same three non-parallel surfaces; The point cloud data correction unit is used to correct the point cloud data collected by each radar according to the installation position of the radar, and obtain the corrected point cloud data corresponding to each radar; a calibration plane determination unit, configured to determine, for each radar, point cloud data corresponding to each of the three non-parallel surfaces from the corresponding corrected point cloud data, and to fit the corresponding calibration plane using the point cloud data of each surface; a calibration parameter determination unit, configured to determine the calibration parameters corresponding to each radar using the obtained calibration plane; A candidate point set updating unit is used to select, for each calibration plane, points from the corresponding candidate point set whose distance to the calibration plane is less than a preset threshold to form a new candidate point set, and to determine whether the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold; an iterative fitting unit, configured to, when a ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is less than a preset threshold, fit a new calibration plane using the new candidate point set, and return to the step of selecting points from the corresponding candidate point set whose distance to the calibration plane is less than the preset threshold; The calibration plane obtaining unit is configured to obtain a calibration plane corresponding to each fitting plane if the ratio of the number of points in the new candidate point set to the number of points in the corresponding candidate point set is not less than a preset threshold.

7. The device according to claim 6, characterized in that Also includes: A correction unit is used to correct the point cloud data collected by each radar according to the installation position of the radar to obtain the corrected point cloud data corresponding to each radar; The calibration unit is used to calibrate the corrected point cloud data of each radar using the calibration parameters corresponding to the radar.

8. A multi-radar data calibration parameter acquisition device, characterized in that: include: memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the multi-radar data calibration parameter acquisition method according to any one of claims 1 to 5.

9. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, each step of the method for acquiring multi-radar data calibration parameters according to any one of claims 1 to 5 is implemented.

Images