A method for evaluating the accuracy of relative pose measurement system of non-cooperative target using LiDAR
Through calibration of lidar and high-precision ground three-dimensional laser scanner and seven-parameter coordinate conversion methods, an evaluation method for lidar non-cooperative target relative posture measurement system was established, and the complexity of lidar system performance evaluation was solved, realizing the realization of the realization of lidar system performance evaluation. High-precision posture measurement and system design support.
Patent Information
- Application Number
- CN202310063557.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-01-11
- Publication Date
- 2025-01-03
- Estimated Expiration
- 2043-01-11
AI Technical Summary
When designing a relative navigation measurement system for lidar for non-cooperative targets, it is difficult to effectively evaluate system performance, especially due to the complexity of sensor data quality, processing algorithms, perspective angles and environmental conditions, and the geometric characteristics of the measured target, resulting in the system There is a lack of standardized approaches to performance evaluation.
By calibrating lidar and high-precision ground three-dimensional laser scanner, a relative relationship matrix is established using the seven-parameter coordinate conversion method, and precision scanning and measurement is carried out in combination with target markers, providing position reference values to evaluate the relative position measurement of lidar Accuracy.
It realizes high-precision posture measurement of the lidar system, provides an evaluation plan for system design and development, and improves the accuracy and reliability of the performance evaluation of the lidar non-cooperative target relative navigation measurement system.
Smart Images
Figure CN115951339B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of relative navigation measurement of space targets, and in particular to an accuracy evaluation method for a laser radar non-cooperative target relative posture measurement system. Background Art
[0002] The process of using LiDAR to estimate the target pose is related to the specific usage scenario and changes with the change of the scenario. If the applicable conditions and application scenarios of the system are not determined first, the system cannot be evaluated. The performance of the LiDAR relative pose measurement system depends on the quality of the sensor data, the processing algorithm, the viewing angle, the environmental conditions, and also on the geometric characteristics of the target being measured in the LiDAR field of view. In order to evaluate its performance, the performance indicators of specific systems are increasingly valued by measuring them through external methods. When designing a LiDAR relative pose measurement system for non-cooperative targets, it is necessary to establish a performance evaluation technical framework for LiDAR systems for relative navigation of non-cooperative targets, and provide evaluation solutions and technical support for the system design and development of LiDAR space non-cooperative target relative navigation measurement systems. Summary of the invention
[0003] The present invention aims to provide a method for evaluating the accuracy of a laser radar non-cooperative target relative position and posture measurement system, and to provide an evaluation scheme and technical support for the system design and development of a laser radar space non-cooperative target relative navigation measurement system.
[0004] In order to achieve the above object, the technical solution of the present invention is as follows: A method for evaluating the accuracy of a laser radar non-cooperative target relative pose measurement system comprises the following steps:
[0005] S1. Calibrate the relative relationship between the laser radar and the high-precision ground 3D laser scanner;
[0006] S2, target relative posture accuracy evaluation test;
[0007] S3. Compare the pose reference value and the LiDAR pose measurement result, and count and analyze their precision and accuracy.
[0008] Furthermore, the calibration method of step S1 is as follows:
[0009] S1.1. Evenly arrange three or more target markers in the overlapping fields of view of the laser radar and the ground-based three-dimensional laser scanner, obtain the three-dimensional point cloud data of the target markers in the laser radar and the ground-based three-dimensional laser scanner, manually extract the target marker point cloud and fit the marker center;
[0010] S1.2. Measure the three-dimensional coordinates P of the marker in the coordinate system of the two devices LRF and P SRF, through the seven-parameter coordinate transformation method, the relative rotation matrix between the two devices is obtained and the translation matrix T.
[0011] Furthermore, the seven-parameter coordinate transformation method model formula of step S1.2 is as follows:
[0012]
[0013] Among them, (X1, Y1, Z1) is the coordinate of the marker point in the laser radar measurement coordinate system, and (X2, Y2, Z2) is the coordinate of the marker point in the ground laser scanner measurement coordinate system; the formula contains three translation parameters (ΔX, ΔY, ΔZ), three rotation parameters (ε x ,ε y ,ε z ) and a scale parameter m, which is a kind of similarity transformation; in order to solve the 7 parameters, at least 3 pairs of common points are required, and the Gaussian least squares method is used to solve it:
[0014]
[0015] T=[ΔX ΔY ΔZ] T .
[0016] Furthermore, the specific steps of step S2 are as follows:
[0017] S2.1. After step S1 is completed, keep the position and posture of the laser radar and the ground 3D laser scanner unchanged. Place the target model on the six-degree-of-freedom movable platform in front of the laser radar line of sight, and affix the special logo sticker for the ground laser 3D scanner;
[0018] S2.2. Design different relative positions and postures of target models. Under each set of posture parameter conditions, use laser radar to perform multiple frame repeated posture measurements on the target model and record the posture measurement results.
[0019] S2.3, use the ground 3D laser scanner to accurately scan the target model landmarks under the current posture conditions, and record the center coordinates of the landmarks; since the 3D coordinates of the landmarks in the target model coordinate system are known, the seven-parameter coordinate transformation method is also used to transform the coordinates between the laser 3D scanner and the target body coordinate system. Solution;
[0020] S2.4, pose reference value calculation; after relative relationship calibration, the above pose results are converted to the laser radar measurement coordinate system, and the pose reference value is
[0021] Compared with the prior art, this solution has the following beneficial effects:
[0022] 1. This solution uses a high-precision, high-density ground laser scanner to establish accurate laser radar exterior orientation elements through calibration field three-dimensional point cloud matching, and then measures the pose of the same target, thereby providing a high-precision target relative pose reference value to achieve accuracy evaluation of the laser radar pose measurement system. The method of this solution is simple to implement and can provide an evaluation solution and technical support for the system design and development of laser radar space non-cooperative target relative navigation measurement system.
[0023] 2. This scheme proposes a cooperative target pose measurement technology based on a high-precision terrestrial laser 3D scanner as a reference for the laser imaging radar pose measurement technology. Special target markers for the laser 3D scanner are arranged on the target surface. The high-precision laser 3D scanner automatically and precisely scans and extracts the center coordinates of the markers. The relative pose parameters of the target can be solved based on the marker coordinates. The relative relationship matrix between the laser radar and the terrestrial 3D laser scanner is combined to unify the measurement benchmark and form an evaluation benchmark value for the relative pose measurement value. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 It is a flow chart of a method for evaluating the accuracy of a laser radar non-cooperative target relative posture measurement system of the present invention;
[0025] Figure 2 is a schematic diagram of a comprehensive test site in this embodiment;
[0026] Figure 3 is a diagram showing the relative position measurement result of the laser radar to the target model in this embodiment;
[0027] Figure 4 3 is a diagram showing the relative posture measurement result of the laser radar to the target model in this embodiment. DETAILED DESCRIPTION
[0028] The present invention is further described in detail below through specific embodiments:
[0029] Example
[0030] The present invention takes a common artificial satellite configuration as an example, and constructs a comprehensive evaluation test field (such as a 64-line laser radar, a ground-based three-dimensional laser scanner, a target model, etc.) Figure 2 As shown), the implementation effect of the present invention is described.
[0031] As attached Figure 1 As shown, a method for evaluating the accuracy of a laser radar non-cooperative target relative pose measurement system includes the following steps:
[0032] S1. First refer to Figure 2To build a comprehensive evaluation test site, calibrate the relative relationship between the laser radar and the high-precision ground 3D laser scanner, and then solve the relative relationship rotation matrix and translation matrix. The calibration method is as follows:
[0033] S1.1. Evenly arrange three or more disc-shaped target markers in the overlapping fields of view of the laser radar and the ground-based three-dimensional laser scanner, obtain the three-dimensional point cloud data of the target markers in the laser radar and the ground-based three-dimensional laser scanner, manually extract the target marker point cloud and fit the marker center.
[0034] S1.2. Measure the three-dimensional coordinates P of the marker in the coordinate system of the two devices LRF and P SRF , through the seven-parameter coordinate transformation method, the relative rotation matrix between the two devices is obtained and the translation matrix T.
[0035] The model formula of the seven-parameter coordinate transformation method in step S1.2 is as follows:
[0036]
[0037] Among them, (X1, Y1, Z1) is the coordinate of the marker point in the laser radar measurement coordinate system, and (X2, Y2, Z2) is the coordinate of the marker point in the ground laser scanner measurement coordinate system; the formula contains three translation parameters (ΔX, ΔY, ΔZ), three rotation parameters (ε x ,ε y ,ε z ) and a scale parameter m, which is a kind of similarity transformation; in order to solve the 7 parameters, at least 3 pairs of common points are required, and the Gaussian least squares method is used to solve it:
[0038]
[0039] Y=[ΔX ΔY ΔZ] T .
[0040] S2. Target relative posture accuracy evaluation test; the specific steps are as follows:
[0041] S2.1. After step S1 is completed, keep the position and posture of the laser radar and the ground 3D laser scanner unchanged. Place the target model on the six-degree-of-freedom movable platform in front of the laser radar line of sight, and paste the special logo sticker of the ground laser 3D scanner.
[0042] S2.2. Design different relative positions and postures of target models. Under each set of posture parameter conditions, use laser radar to perform multiple frame repeated posture measurements on the target model and record the posture measurement results. In this embodiment, the target is placed about 13 meters in front of the laser radar line of sight. The laser radar is first used to perform 80 frames of repeated posture measurement of the target, and 80 sets of posture measurement results are recorded.
[0043] S2.3, use the ground 3D laser scanner to accurately scan the target model landmarks under the current posture conditions, and record the center coordinates of the landmarks; since the 3D coordinates of the landmarks in the target model coordinate system are known, the seven-parameter coordinate transformation method is also used to transform the coordinates between the laser 3D scanner and the target body coordinate system. Please solve.
[0044] S2.4, pose reference value calculation; after relative relationship calibration, the above pose results are converted to the laser radar measurement coordinate system, and the pose reference value is In this embodiment, the position reference values are X=12.873m, Y=0.035m, Z=-0.053m, and the attitude reference values are X=93.847°, Y=-0.908°, Z=-1.743°.
[0045] S3, perform posture accuracy evaluation, and measure the posture of 80 frames of LiDAR Compare with the reference value R and calculate its precision and accuracy. Figure 3 and 4 As shown, Figure 3 Among them, (1) is the X-axis position result, (2) is the Y-axis position result, and (3) is the Z-axis position result; Figure 4 (2) is the result of the X-axis attitude angle, (3) is the result of the Y-axis attitude angle, and (4) is the result of the Z-axis attitude angle. Compared with the reference value, the relative position measurement accuracy and precision of the laser radar in this embodiment are 1.8cm±1mm for the X-axis, 0.5cm±2mm for the Y-axis, and 0.5cm±4mm for the Z-axis. The relative attitude measurement accuracy and precision are 0.243°±0.060° for the X-axis, 0.156°±0.074° for the Y-axis, and 0.810°±0.086° for the Z-axis.
[0046] The above are only embodiments of the present invention, and the common knowledge such as the known specific structures and / or characteristics in the scheme are not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.
Claims
1. A method for evaluating the accuracy of a laser radar non-cooperative target relative pose measurement system, characterized in that: The steps include: S1. Calibrate the relative relationship between the laser radar and the high-precision ground 3D laser scanner. The calibration method is as follows: S1.
1. Evenly arrange three or more target markers in the overlapping fields of view of the laser radar and the ground-based three-dimensional laser scanner, obtain the three-dimensional point cloud data of the target markers in the laser radar and the ground-based three-dimensional laser scanner, manually extract the target marker point cloud and fit the marker center; S1.
2. Measure the three-dimensional coordinates of the marker in the coordinate system of the two devices and , through the seven-parameter coordinate transformation method, the relative rotation matrix between the two devices is obtained and translation matrix ; S2. Target relative posture accuracy evaluation test, including: S2.
1. After step S1 is completed, keep the position and posture of the laser radar and the ground 3D laser scanner unchanged; place the target model on a six-degree-of-freedom movable platform in front of the laser radar line of sight, and affix a special logo sticker for the ground 3D laser scanner; S2.
2. Design different relative positions and postures of target models. Under each set of posture parameter conditions, use laser radar to perform multiple frame repeated posture measurements on the target model and record the posture measurement results. ; S2.
3. Use a terrestrial 3D laser scanner to precisely scan the terrestrial 3D laser scanner special label of the target model under the current posture condition, and record the center coordinates of the terrestrial 3D laser scanner special label of the target model; since the 3D coordinates of the terrestrial 3D laser scanner special label of the target model in the target model coordinate system are known, the seven-parameter coordinate transformation method is also used to transform the coordinate parameters between the terrestrial 3D laser scanner and the target body coordinate system. Solution; S2.4, calculation of the reference value of the posture; after calibration through relative relationship, the above coordinate conversion parameters Converted to the laser radar measurement coordinate system, the pose reference value is ; S3. Compare the pose reference value and the LiDAR pose measurement result, and count and analyze their precision and accuracy.
2. The method for evaluating the accuracy of a laser radar non-cooperative target relative pose measurement system according to claim 1, characterized in that: The model formula of the seven-parameter coordinate transformation method in step S1.2 is as follows: ； in, is the coordinate of the marker point in the laser radar measurement coordinate system, is the coordinate of the marker point in the ground 3D laser scanner measurement coordinate system; the formula contains three translation parameters , three rotation parameters and a scale parameter , which is a kind of similarity transformation; in order to solve the 7 parameters, at least 3 pairs of common points are required, and the Gaussian least squares method is used to solve it: 。