An adjustment method and system for calibrating an external parameter, an electronic device, and a medium

Abstract

The application discloses a kind of calibration external parameter adjustment method, system, electronic equipment and medium, the method includes: with the actual GPS position of target point cloud data as standard, the actual azimuth difference mean produced when the target point cloud data is converted from radar coordinate system to geodetic coordinate system based on calibration external parameter is determined;The actual azimuth difference mean is used to characterize the overall deviation of the target point cloud data from radar coordinate system conversion to the geodetic coordinate system relative to the actual GPS position;The calibration external parameter includes three-axis rotation vector and three-axis translation vector;The three-axis translation vector is corrected based on the actual azimuth difference mean;After the three-axis translation vector correction, the three-axis rotation vector is iteratively adjusted according to the set step length and the distance variance of the target point cloud data is calculated correspondingly, until the distance variance is within the set threshold.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to a method, system, electronic device and medium for adjusting calibrated external parameters. Background Technology

[0002] With the rapid development of intelligent transportation, the use of multi-line LiDAR, cameras, millimeter-wave radar and other sensing devices for vehicle-road cooperative sensing has become the norm. These devices can provide accurate roadside sensing data, thereby improving the safety level of intelligent transportation.

[0003] As is well known, when radar tracks a target, the point cloud data collected by radar is three-dimensional data in the radar coordinate system. Therefore, when constructing a point cloud map, it is necessary to first calibrate the relative pose parameters of the radar coordinate system with respect to the geodetic coordinate system (also known as calibrating extrinsic parameters), and then project the point cloud data into the geodetic coordinate system to construct the point cloud map.

[0004] Existing extrinsic parameter calibration typically involves mapping the same calibration object's GPS coordinates in the world coordinate system to its point cloud coordinates in the radar coordinate system. While this method is simple and quick, it doesn't account for potential positional deviations caused by environmental factors during actual radar operation. When the radar's position deviates, the tracking GPS coordinates obtained through extrinsic parameter conversion will also experience mapping errors, resulting in discrepancies with the actual GPS coordinates.

[0005] Therefore, a more accurate method for calibrating and adjusting extrinsic parameters is needed to solve the mapping error problem in calibrating extrinsic parameters. Summary of the Invention

[0006] This invention provides a method, system, electronic device, and medium for adjusting calibrated extrinsic parameters, in order to solve or partially solve the problem of mapping error in calibrated extrinsic parameters caused by the influence of the external environment.

[0007] To address the aforementioned technical problems, a first aspect of the present invention discloses a method for adjusting calibrated extrinsic parameters, the method comprising:

[0008] Using the actual GPS position of the target point cloud data as a standard, the mean actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on calibration extrinsic parameters is determined; the mean actual azimuth difference is used to characterize the overall offset of the target point cloud data relative to the actual GPS position after the transformation from the radar coordinate system to the geodetic coordinate system; the calibration extrinsic parameters include a three-axis rotation vector and a three-axis translation vector.

[0009] The three-axis translation vector is corrected based on the average actual azimuth difference.

[0010] After the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to a set step size, and the distance variance of the target point cloud data is calculated accordingly, until the distance variance is within a set threshold.

[0011] Optionally, before determining the average actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on the actual GPS position of the target point cloud data, the method further includes:

[0012] The target point cloud data is determined from the point cloud data in the radar coordinate system;

[0013] The actual GPS coordinates of the target object corresponding to the target point cloud data are obtained by measuring with a positioning device.

[0014] Using the calibration extrinsic parameters, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain the GPS theoretical coordinate values ​​of the target point cloud data.

[0015] Optionally, determining the average actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on calibration extrinsic parameters specifically includes:

[0016] Using the actual GPS coordinates and theoretical GPS coordinates of the target point cloud data, determine the actual deviation distance of the target point cloud data;

[0017] The actual deviation distance is projected onto latitude and longitude directions and the mean is calculated to obtain the mean actual azimuth difference of the target point cloud data in the latitude and longitude directions.

[0018] Optionally, the average actual azimuth difference includes the actual deviation distance and the actual deviation direction;

[0019] The step of correcting the three-axis translation vector using the mean of the actual azimuth difference specifically includes:

[0020] The three-axis translation vector is shifted in the opposite direction to the actual deviation direction; the displacement of the three-axis translation vector depends on the actual deviation distance.

[0021] Optionally, after the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to a set step size, and the distance variance of the target point cloud data is calculated accordingly, until the distance variance is within a set threshold. This specifically includes:

[0022] After the three-axis translation vector is corrected, the roll angle step size and roll boundary are set with the roll angle of the three-axis rotation vector as the correction target;

[0023] Under the constraint of the tumbling boundary, the tumbling angle is iteratively adjusted according to the tumbling angle step size, and the distance variance of the target point cloud data is calculated accordingly.

[0024] Determine whether the calculated distance variance is within the set threshold.

[0025] If so, stop iterating;

[0026] If not, continue iterative adjustments until the distance variance is within the set threshold.

[0027] Optionally, calculating the distance variance of the target point cloud data specifically includes:

[0028] According to the modified roll angle, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain GPS mapped coordinate values;

[0029] The mapping deviation distance is determined using the actual GPS coordinates and the mapped GPS coordinates.

[0030] The mean value of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions is obtained by projecting the mapping deviation distance onto the latitude and longitude directions respectively and calculating the mean value.

[0031] The distance variance is calculated by using the mean of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions.

[0032] Optionally, after iteratively adjusting the three-axis rotation vector according to a set step size and calculating the distance variance of the target point cloud data accordingly, until the distance variance is within a set threshold, the method further includes:

[0033] The three-axis translation vector is iteratively corrected based on the distance variance within the set threshold.

[0034] Each time the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to the set step size, and the distance variance of the target point cloud data is calculated accordingly, until the fluctuation of the distance variance within the set threshold is within the fluctuation range.

[0035] A second aspect of the present invention discloses a calibration extrinsic parameter adjustment system, the system comprising:

[0036] The determination module is used to determine the average actual azimuth difference of the target point cloud data when it is transformed from the radar coordinate system to the geodetic coordinate system based on the actual GPS position of the target point cloud data. The average actual azimuth difference is used to characterize the overall offset of the target point cloud data relative to the actual GPS position after the transformation from the radar coordinate system to the geodetic coordinate system. The calibration azimuth parameters include a three-axis rotation vector and a three-axis translation vector.

[0037] The correction module is used to correct the three-axis translation vector using the mean of the actual azimuth difference;

[0038] The adjustment module is used to iteratively adjust the three-axis rotation vector according to a set step size after the three-axis translation vector is corrected, and to calculate the distance variance of the target point cloud data accordingly, until the distance variance is within a set threshold.

[0039] A third aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the above-described method.

[0040] A fourth aspect of the present invention discloses an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described method.

[0041] Through one or more technical solutions of the present invention, the present invention has the following beneficial effects or advantages:

[0042] The scheme disclosed in this invention uses the actual GPS position of the target point cloud data as a standard to determine the average actual azimuth difference of the target point cloud data. Since the average actual azimuth difference can characterize the overall offset of the target point cloud data relative to the actual GPS position after transformation from the radar coordinate system to the geodetic coordinate system, the average actual azimuth difference is used as a standard to correct the three-axis translation vector in the calibration extrinsic parameters. After the three-axis translation vector is corrected, the three-axis rotation vector in the calibration extrinsic parameters is iteratively adjusted according to a set step size. Therefore, this scheme utilizes the average actual azimuth difference to account for the positional deviation of the radar coordinate system caused by external environmental influences. By referencing the average actual azimuth difference to fine-tune the calibration extrinsic parameters, it can eliminate the mapping error caused by external environmental influences in the calibration extrinsic parameters, thereby improving the calibration accuracy of the calibration extrinsic parameters in actual use.

[0043] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0044] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0045] In the attached diagram:

[0046] Figure 1 A schematic flowchart of a method for adjusting calibration extrinsic parameters according to an embodiment of the present invention is shown;

[0047] Figure 2 A schematic diagram of the iterative cyclic adjustment process of a calibration extrinsic parameter adjustment method according to an embodiment of the present invention is shown;

[0048] Figure 3 A schematic diagram of the iterative adjustment process of the three-axis translation vector and the three-axis rotation vector according to an embodiment of the present invention is shown;

[0049] Figure 4 A schematic diagram of a calibration extrinsic parameter adjustment system according to an embodiment of the present invention is shown;

[0050] Figure 5 A schematic diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0051] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0052] This invention discloses a method for adjusting calibration extrinsic parameters, which is mainly used to adjust the relative pose parameters of the radar coordinate system and the geodetic coordinate system. This method is used to fine-tune the calibration extrinsic parameters, thereby achieving higher calibration accuracy.

[0053] Figure 1 A flowchart illustrating the method for adjusting calibration extrinsic parameters provided in the embodiments of this specification is shown. See also... Figure 1 The method for adjusting the calibration extrinsic parameters provided in the embodiments of this specification includes the following steps S101 to S104:

[0054] S101, using the actual GPS position of the target point cloud data as the standard, determines the average actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on the calibration extrinsic parameters.

[0055] The target point cloud data is a set of point cloud data extracted from the point cloud data in the radar coordinate system. The point cloud data acquired by the radar is three-dimensional data presented in the radar coordinate system, with the radar transmission center as the origin, the positive Y-axis directly in front of the transmission surface as the positive X-axis, the positive right side as the positive Z-axis, and the positive top as the positive Z-axis. The geodetic coordinate system is a coordinate system established in geodesy with a reference ellipsoid as the datum. This invention uses a general positioning device as the reference for the geodetic coordinate system, such as an RTK (Real-time kinematic) positioning device. The pose data output by the RTK positioning device is its relative pose with respect to the origin of the geodetic coordinate system. Therefore, after determining the relative pose between the RTK positioning device and the radar, the relative pose of the origin of the radar coordinate system with respect to the origin of the geodetic coordinate system can be obtained.

[0056] The calibration extrinsic parameters are used to characterize the relative pose parameters of the radar coordinate system and the geodetic coordinate system, including the three-axis rotation vector r and the three-axis translation vector t. The three-axis rotation vector r characterizes the three-axis rotation angles of the radar coordinate system relative to the geodetic coordinate system, including: pitch angle rx, yaw angle ry, and roll angle rz; the three-axis translation vector t characterizes the three-axis translation distances of the radar coordinate system relative to the geodetic coordinate system, including: X-axis translation distance tx, Y-axis translation distance ty, and Z-axis translation distance tz.

[0057] To achieve the goal of fine-tuning the extrinsic parameters, this embodiment uses the average actual azimuth difference as the reference standard for fine-tuning. The average actual azimuth difference is used to characterize the overall offset of the target point cloud data relative to the actual GPS position after the radar coordinate system is transformed to the geodetic coordinate system. It takes into account both the actual deviation distance and the actual deviation direction of the target point cloud data, and therefore can accurately characterize the actual offset of the target object corresponding to the target point cloud data.

[0058] Furthermore, the average actual azimuth difference is obtained by projecting the relative distance between the actual GPS coordinates and the theoretical GPS coordinates of the target point cloud data onto the latitude and longitude directions. Therefore, it is necessary to obtain the actual GPS coordinates and the theoretical GPS coordinates beforehand. Specifically, the target point cloud data is determined from the point cloud data in the radar coordinate system; the actual GPS coordinates of the target object corresponding to the target point cloud data are measured using positioning equipment; and then, using calibration extrinsic parameters, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain the theoretical GPS coordinates of the target point cloud data. The actual GPS coordinates characterize the actual GPS position of the target point cloud data in the geodetic coordinate system. The theoretical GPS coordinates characterize the mapped GPS position of the target point cloud data from the radar coordinate system to the geodetic coordinate system.

[0059] In the process of calculating the average actual azimuth difference of the target point cloud data, the actual deviation distance of the target point cloud data is determined by using the actual GPS coordinates and theoretical GPS coordinates of the target point cloud data. The actual deviation distance is then projected onto the latitude and longitude directions and the average is calculated to obtain the average actual azimuth difference of the target point cloud data in the latitude and longitude directions.

[0060] Specifically, in latitude and longitude directions, due east is the positive latitude direction, due west is the negative latitude direction, due north is the positive longitude direction, and due south is the negative longitude direction. The actual deviation distance is the relative distance *d* between the theoretical GPS coordinates and the actual GPS coordinates of the target point cloud data. This relative distance *d* is projected onto both the longitude and latitude directions and the mean is calculated to obtain the average actual azimuth difference in the latitude and longitude directions. This includes the average actual azimuth difference *dx* in the longitude direction and the average actual azimuth difference *dy* in the latitude direction.

[0061] During the calculation process, a set number of maximum and minimum values ​​are first removed from the target point cloud data. Then, the average actual azimuth difference is determined based on the remaining target point cloud data. For example, if the actual deviation distance of the remaining target point cloud data is (d1, d2, ..., d...), then... n ), where n represents the number of remaining target point cloud data, then the average actual azimuth difference of the remaining target point cloud data in the longitude direction is: Where, d ix This represents the average actual azimuth difference of the i-th reference point in the longitude direction. The average actual azimuth difference of the remaining target point cloud data in the latitude direction is:

[0062] In this system, dx and dy have positive and negative values, with the sign indicating the actual direction of deviation of the radar coordinate system from the geodetic coordinate system in terms of latitude and longitude. For example, a negative dx indicates that the longitude of the target point cloud data is too low; a positive dx indicates that the longitude of the target point cloud data is too high; a negative dy indicates that the latitude of the target point cloud data is too low; and a positive dy indicates that the latitude of the target point cloud data is too high. The specific values ​​of dx and dy represent the actual deviation distance.

[0063] S102, using the average value of the actual azimuth difference to correct the three-axis translation vector.

[0064] The actual azimuth difference mean includes the actual deviation distance and actual deviation direction of the target point cloud data. Therefore, the three-axis translation vector is shifted in the opposite direction according to the actual deviation direction to correct the deviation direction of the three-axis translation vector. The displacement of the three-axis translation vector depends on the actual deviation distance.

[0065] Specifically, according to the formula Correct the three-axis translation vector; where tx' represents the corrected translation distance along the X-axis and ty' represents the corrected translation distance along the Y-axis. For example, if dx = 4.2 and dy = 4.2, then shifting the X-axis translation distance tx westward and the Y-axis translation distance ty southward by 4.2 will correct the three-axis translation vector.

[0066] In traditional three-axis translation vector adjustment methods, only the relative deviation distance of the target point cloud data is considered, without taking into account the direction of the deviation. However, this proposed method comprehensively considers both the actual offset distance and the actual deviation direction of the target point cloud data when mapped to the geodetic coordinate system. This allows for the introduction of adjustment orientation into the three-axis translation vector adjustment, providing directional information for fine-tuning the calibration extrinsic parameters. Consequently, it achieves more accurate calibration extrinsic parameters than existing methods.

[0067] It is worth noting that S102 to S103 are executed in a specific order. First, the three-axis translation vector needs to be corrected to ensure the origins of the radar coordinate system and the geodetic coordinate system coincide as much as possible. Then, the three-axis rotation vector is corrected to improve its adjustment accuracy. Furthermore, S102 to S103 is an iterative process; by continuously adjusting the three-axis translation and rotation vectors, the adjustment accuracy of the calibration extrinsic parameters is maximized.

[0068] S103, after the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to the set step size and the distance variance of the target point cloud data is calculated accordingly until the distance variance is within the set threshold.

[0069] Specifically, after correcting the three-axis translation vector, the pitch and yaw angles of the three-axis rotation vector are kept in the same plane as the geodetic coordinate system. The roll angle of the three-axis rotation vector is used as the correction target to set the roll angle step size and roll boundary. For example, the roll angle step size is set to 0.1°, and the roll boundary is set to ±10°. It is worth noting that the roll angle step size and roll boundary can be set according to empirical values ​​or with reference to the deviation direction of the target point cloud data relative to the geodetic coordinate system. In addition, since the three-axis translation vector and the three-axis rotation vector will continuously iterate, after each correction of the three-axis translation vector, the roll angle step size and roll boundary can also be set with reference to the distance variance used in that correction. In practical applications, the roll angle step size and roll boundary can be set in any of the above methods without limitation.

[0070] Under the constraint of the roll boundary, the roll angle is iteratively adjusted according to the roll angle step size, and the distance variance of the target point cloud data is calculated accordingly.

[0071] In calculating the range variance, it is necessary to first calculate the mean of the mapping azimuth difference of the target point cloud data, and then calculate the range variance. Specifically, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system according to the modified roll angle, obtaining the GPS mapped coordinate values. Using the actual GPS coordinate values ​​and the GPS mapped coordinate values, the mapping deviation distance d' is determined. The mapping deviation distance d' is then projected onto the latitude and longitude directions and its mean is calculated to obtain the mean of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions, including: the mean mapping azimuth difference dx' in the longitude direction and the mean mapping azimuth difference dy' in the latitude direction. Similarly, dx' and dy' have positive and negative values, with the positive and negative signs indicating the direction of the mapping deviation of the radar coordinate system relative to the geodetic coordinate system in the latitude and longitude directions. For example, a negative dx' indicates that the longitude of the target point cloud data is too small; a positive dx' indicates that the longitude of the target point cloud data is too large; a negative dy' indicates that the latitude of the target point cloud data is too small; a positive dy' indicates that the latitude of the target point cloud data is too large. The specific values ​​of dx' and dy' represent the mapping deviation distance. The distance variance is calculated using the mean of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions. The calculation method for the distance variance is as follows: Where m represents the number of target point cloud data, m≥n, that is: m can take all target point cloud data, or it can take the remaining reference points after excluding reference points with large errors.

[0072] It is worth noting that after each iteration of adjusting the roll angle, the distance variance must be calculated, and it must be determined whether the calculated distance variance is within a set threshold. The setting threshold can be further restricted to minimizing the distance variance under the constraints of the roll boundary. If so, it indicates that the three-axis rotation vector has been optimally adjusted under the constraints of the roll boundary, and the iteration stops. If not, the iteration continues until the calculated distance variance reaches within the set threshold.

[0073] The above describes the iterative adjustment process of the three-axis rotation vector, which allows the three-axis rotation vector to be adjusted to its optimal state within the constraints of the tumble boundary. To avoid the three-axis rotation vector getting trapped in local optima, please refer to... Figure 2 This is a flowchart illustrating the iterative adjustment process for the calibration of extrinsic parameters.

[0074] After obtaining the distance variance within the set threshold, execute S103 to determine whether the fluctuation of the distance variance within the set threshold is within the fluctuation range. If the fluctuation of the distance variance within the set threshold is within the fluctuation range, it indicates that the distance variance tends to stabilize within the set threshold, and the three-axis translation vector and the three-axis rotation vector simultaneously reach the global optimum, then stop the iteration.

[0075] If the fluctuation range is exceeded, the three-axis translation vector is iteratively corrected. Specifically, S104 is executed, replacing the actual azimuth difference mean with the mapped azimuth difference mean corresponding to the distance variance within the set threshold, and then proceeding to S102 to execute the step of correcting the three-axis translation vector.

[0076] The mean of the mapping azimuth difference includes the mapping deviation distance and the mapping deviation direction. During the correction of the three-axis translation vector, the three-axis translation vector is translated in the opposite direction according to the mapping deviation direction; the displacement of the three-axis translation vector depends on the mapping deviation distance.

[0077] It is worth noting that after each correction of the three-axis translation vector, the three-axis rotation vector must be iteratively adjusted according to a set step size, and the distance variance of the target point cloud data must be calculated accordingly, until the fluctuation of the distance variance within a set threshold is within the fluctuation range. At this point, the roll angle step size and roll boundary are set with the roll angle of the three-axis rotation vector as the correction target. For example, the roll angle step size and roll boundary are set with reference to the distance variance of the three-axis translation vector in this correction. The method for adjusting the distance variance can refer to a similar scheme described above, and will not be repeated here.

[0078] To facilitate the explanation and interpretation of the iterative adjustment process of the three-axis translation vector and the three-axis rotation vector, please refer to the following: Figure 3 Specific examples will be used to illustrate this.

[0079] S301, using the actual GPS position of the target point cloud data as the standard, determine the average actual azimuth difference (dx, dy) of the target point cloud data.

[0080] The mean actual azimuth difference (dx, dy) of the target point cloud data is based on the initial values ​​of the calibration extrinsic parameters: the three-axis translation vector (tx, ty, tz) and the three-axis rotation vector (rx, ry, rz), which are generated when transforming from the radar coordinate system to the geodetic coordinate system.

[0081] S302 uses the actual azimuth difference mean (dx, dy) to correct the three-axis translation vector.

[0082] Specifically, using formulas Correct the three-axis translation vector.

[0083] S303, set the roll angle step size in the three-axis rotation vector to 0.1°, and set the roll boundary to ±10°.

[0084] S304, the roll angle rz in the three-axis rotation vector is changed by 0.1°.

[0085] S305, determine if the roll angle rz is within [-10°, +10°]. If not, stop iteration. If yes, execute S306.

[0086] S306, determine the GPS mapping coordinates of the target point cloud data, and calculate the mean of the mapping azimuth difference and the variance of the distance.

[0087] S307: Determine if the distance variance is the minimum distance variance within the roll boundary [-10°, +10°]. If not, return to S304. If yes, proceed to S308.

[0088] S308, determine whether the fluctuation of the distance variance is within the set threshold. If yes, stop the iteration. If no, proceed to S309.

[0089] S309, replace the actual azimuth difference with the mapped azimuth difference mean, and then proceed to S302 for execution.

[0090] The above describes the iterative adjustment process for the three-axis translation vector and the three-axis rotation vector. This iterative adjustment process continues until the fluctuation of the distance variance within a set threshold is within the acceptable range, or until the required number of iterations is met. After multiple iterative adjustments, more accurate calibration parameters can be obtained than those of existing calibration methods.

[0091] The above describes all implementation schemes for adjusting the calibration extrinsic parameters of the present invention. In the scheme of the present invention, the actual GPS position of the target point cloud data is used as the standard to determine the average actual azimuth difference of the target point cloud data. Since the average actual azimuth difference is used to characterize the overall offset of the target point cloud data relative to the actual GPS position after being transformed from the radar coordinate system to the geodetic coordinate system, the average actual azimuth difference is used as the standard to correct the three-axis translation vector in the calibration extrinsic parameters. After the three-axis translation vector is corrected, the three-axis rotation vector in the calibration extrinsic parameters is iteratively adjusted according to a set step size. It can be seen that this scheme takes into account the position deviation of the radar coordinate system caused by the influence of the external environment by using the average actual azimuth difference. By referring to the average actual azimuth difference to fine-tune the calibration extrinsic parameters, the mapping error of the calibration extrinsic parameters caused by the influence of the external environment can be eliminated, thereby improving the calibration accuracy of the calibration extrinsic parameters in actual use.

[0092] Based on the same inventive concept as in the foregoing embodiments, this invention also discloses a calibration extrinsic parameter adjustment system, see below. Figure 4 The system includes:

[0093] The determination module 401 is used to determine the average actual azimuth difference of the target point cloud data when it is transformed from the radar coordinate system to the geodetic coordinate system based on the actual GPS position of the target point cloud data, using the actual GPS position of the target point cloud data as the standard; the average actual azimuth difference is used to characterize the overall offset of the target point cloud data relative to the actual GPS position after the transformation from the radar coordinate system to the geodetic coordinate system; the calibration azimuth parameters include a three-axis rotation vector and a three-axis translation vector.

[0094] Correction module 402 is used to correct the three-axis translation vector using the mean of the actual azimuth difference;

[0095] The adjustment module 403 is used to iteratively adjust the three-axis rotation vector according to a set step size after the three-axis translation vector is corrected, and calculate the distance variance of the target point cloud data accordingly, until the distance variance is within a set threshold.

[0096] In one optional implementation, the system further includes: an acquisition module, configured to:

[0097] The target point cloud data is determined from the point cloud data in the radar coordinate system;

[0098] The actual GPS coordinates of the target object corresponding to the target point cloud data are obtained by measuring with a positioning device.

[0099] Using the calibration extrinsic parameters, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain the GPS theoretical coordinate values ​​of the target point cloud data.

[0100] In one optional implementation, the determining module 401 is specifically used for:

[0101] Using the actual GPS coordinates and theoretical GPS coordinates of the target point cloud data, determine the actual deviation distance of the target point cloud data;

[0102] The actual deviation distance is projected onto latitude and longitude directions and the mean is calculated to obtain the mean actual azimuth difference of the target point cloud data in the latitude and longitude directions.

[0103] In one optional implementation, the average actual azimuth difference includes the actual deviation distance and the actual deviation direction;

[0104] The correction module 402 is specifically used for:

[0105] The three-axis translation vector is shifted in the opposite direction to the actual deviation direction; the displacement of the three-axis translation vector depends on the actual deviation distance.

[0106] In one optional implementation, the adjustment module 403 is specifically used for:

[0107] After the three-axis translation vector is corrected, the roll angle step size and roll boundary are set with the roll angle of the three-axis rotation vector as the correction target;

[0108] Under the constraint of the tumbling boundary, the tumbling angle is iteratively adjusted according to the tumbling angle step size, and the distance variance of the target point cloud data is calculated accordingly.

[0109] Determine whether the calculated distance variance is within the set threshold.

[0110] If so, stop iterating;

[0111] If not, continue iterative adjustments until the distance is within the set threshold.

[0112] In one optional implementation, the adjustment module 403 is specifically used for:

[0113] According to the modified roll angle, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain GPS mapped coordinate values;

[0114] The mapping deviation distance is determined using the actual GPS coordinates and the mapped GPS coordinates.

[0115] The mean value of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions is obtained by projecting the mapping deviation distance onto the latitude and longitude directions respectively and calculating the mean value.

[0116] The distance variance is calculated using the mean of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions.

[0117] In an optional implementation, the correction module 402 is further configured to: iteratively correct the three-axis translation vector based on the distance variance within the set threshold.

[0118] The adjustment module 403 is also used to iteratively adjust the three-axis rotation vector according to the set step size and calculate the distance variance of the target point cloud data each time the three-axis translation vector is corrected, until the fluctuation of the distance variance within the set threshold is within the fluctuation range.

[0119] It should be noted that the specific methods by which each module performs operations in the calibration extrinsic parameter adjustment system provided in the embodiments of this specification have been described in detail in the above method embodiments. The specific implementation process can be referred to the above method embodiments, and will not be described in detail here.

[0120] Based on the same inventive concept as in the foregoing embodiments, this embodiment of the invention also discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0121] Based on the same inventive concept as in the foregoing embodiments, this embodiment of the invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0122] like Figure 5 As shown, the electronic device includes a memory 504, one or more processors 502, and a computer program stored in the memory 504 and executable on the processor 502. When the processor 502 executes the program, it implements the steps of any embodiment of the road object height detection method provided in the first or second aspect above. For example, the electronic device may be a device with data processing capabilities, such as an edge computing device, a personal computer, a tablet computer, or a server.

[0123] Among them, Figure 5 In this document, a bus architecture (represented by bus 500) is used. Bus 500 may include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 502 and memory represented by memory 504. Bus 500 may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 505 provides an interface between bus 500 and receiver 501 and transmitter 503. Receiver 501 and transmitter 503 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 502 is responsible for managing bus 500 and general processing, while memory 504 can be used to store data used by processor 502 during operation.

[0124] Understandable Figure 5 The structure shown is for illustrative purposes only. The electronic device provided in the embodiments of this specification may also include components that are more advanced than those shown in the original text. Figure 5 The more or fewer components shown, or having the same Figure 5 The different configurations shown. Figure 5 The components shown can be implemented using hardware, software, or a combination thereof.

[0125] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0126] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 Devices that specify the functions in one or more boxes.

[0127] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction device, which is implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0128] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0129] Although preferred embodiments have been described in this specification, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this specification.

[0130] Obviously, those skilled in the art can make various modifications and variations to this specification without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims and their equivalents, this specification is also intended to include such modifications and variations.

Claims

1. A method for adjusting calibrated extrinsic parameters, characterized in that, The method includes: Using the actual GPS position of the target point cloud data as a standard, the mean value of the actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on calibration extrinsic parameters is determined. Specifically, this includes: determining the actual deviation distance of the target point cloud data using the actual GPS coordinate values ​​and theoretical GPS coordinate values; projecting the actual deviation distance onto latitude and longitude directions and calculating the mean value to obtain the mean value of the actual azimuth difference of the target point cloud data in those directions; the mean value of the actual azimuth difference characterizes the overall offset of the target point cloud data relative to the actual GPS position after transformation from the radar coordinate system to the geodetic coordinate system; the mean value of the actual azimuth difference includes the actual deviation distance and the actual deviation direction; the calibration extrinsic parameters include a three-axis rotation vector and a three-axis translation vector. Correcting the three-axis translation vector based on the actual azimuth difference mean specifically includes: translating the three-axis translation vector in the opposite direction according to the actual deviation direction; the displacement of the three-axis translation vector depends on the actual deviation distance; After the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to a set step size, and the distance variance of the target point cloud data is calculated accordingly, until the distance variance is within a set threshold. Specifically, this includes: after the three-axis translation vector is corrected, setting a roll angle step size and a roll boundary with the roll angle of the three-axis rotation vector as the correction target; under the constraint of the roll boundary, iteratively adjusting the roll angle according to the roll angle step size, and calculating the distance variance of the target point cloud data accordingly; determining whether the calculated distance variance is within the set threshold; if yes, stopping the iteration; if no, continuing the iterative adjustment until the distance variance is within the set threshold.

2. The method as described in claim 1, characterized in that, Before determining the average actual azimuth difference generated when the target point cloud data is transformed from the radar coordinate system to the geodetic coordinate system based on the actual GPS position of the target point cloud data, the method further includes: The target point cloud data is determined from the point cloud data in the radar coordinate system; The actual GPS coordinates of the target object corresponding to the target point cloud data are obtained by measuring with a positioning device. Using the calibration extrinsic parameters, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain the GPS theoretical coordinate values ​​of the target point cloud data.

3. The method as described in claim 1, characterized in that, The calculation of the distance variance of the target point cloud data specifically includes: According to the modified roll angle, the target point cloud data is mapped from the radar coordinate system to the geodetic coordinate system to obtain GPS mapped coordinate values; The mapping deviation distance is determined using the actual GPS coordinates and the mapped GPS coordinates. The mean value of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions is obtained by projecting the mapping deviation distance onto the latitude and longitude directions respectively and calculating the mean value. The distance variance is calculated using the mean of the mapping azimuth difference of the target point cloud data in the latitude and longitude directions.

4. The method as described in claim 3, characterized in that, After iteratively adjusting the three-axis rotation vector according to a set step size and calculating the distance variance of the target point cloud data accordingly, until the distance variance is within a set threshold, the method further includes: The three-axis translation vector is iteratively corrected based on the distance variance within the set threshold. Each time the three-axis translation vector is corrected, the three-axis rotation vector is iteratively adjusted according to the set step size, and the distance variance of the target point cloud data is calculated accordingly, until the fluctuation of the distance variance within the set threshold is within the fluctuation range.

5. A calibration system for adjusting extrinsic parameters, characterized in that, The system includes: The determination module is used to determine the average actual azimuth difference of the target point cloud data when it is transformed from the radar coordinate system to the geodetic coordinate system based on the actual GPS position of the target point cloud data. Specifically, this includes: determining the actual deviation distance of the target point cloud data using the actual GPS coordinate values ​​and theoretical GPS coordinate values; projecting the actual deviation distance onto latitude and longitude directions and calculating the average value to obtain the average actual azimuth difference of the target point cloud data in those directions; the average actual azimuth difference characterizes the overall offset of the target point cloud data relative to the actual GPS position after transformation from the radar coordinate system to the geodetic coordinate system; the average actual azimuth difference includes the actual deviation distance and the actual deviation direction; the calibration extrinsic parameters include a three-axis rotation vector and a three-axis translation vector. The correction module is used to correct the three-axis translation vector using the average actual azimuth difference, specifically including: translating the three-axis translation vector in the opposite direction according to the actual deviation direction; the displacement of the three-axis translation vector depends on the actual deviation distance; The adjustment module is used to iteratively adjust the three-axis rotation vector according to a set step size and calculate the distance variance of the target point cloud data after the three-axis translation vector is corrected, until the distance variance is within a set threshold. Specifically, it includes: after the three-axis translation vector is corrected, setting a roll angle step size and a roll boundary with the roll angle of the three-axis rotation vector as the correction target; under the constraint of the roll boundary, iteratively adjusting the roll angle according to the roll angle step size and calculating the distance variance of the target point cloud data accordingly; determining whether the calculated distance variance is within the set threshold; if yes, stopping the iteration; if no, continuing the iterative adjustment until the distance variance is within the set threshold.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-4.

7. 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 program, it implements the steps of the method according to any one of claims 1-4.

Images