Single-line and multi-line lidar joint calibration method and system based on straight line features

Abstract

The present invention discloses a method and system for joint calibration of single-line and multi-line laser radars based on straight line features. The method comprises: performing preliminary cropping on the multi-line laser radar point cloud, fitting two planes, calculating the projections of the two planes in the Xoy plane of the multi-line laser radar coordinate system to obtain two straight lines, and then adding a number of points along the z-axis; performing preliminary cropping on the single-line laser radar point cloud, fitting two straight lines, and screening the single-line laser radar point cloud so that the two straight lines in the single-line laser radar point cloud are equal in length to the two straight lines in the multi-line laser radar point cloud, and then adding a number of points along the z-axis; aligning the single-line laser radar point cloud with the multi-line laser radar point cloud to obtain the relative pose of the single-line laser radar and the multi-line laser radar, thereby completing the calibration. The present invention has the advantages of low requirements for data acquisition scenarios, simple operation, and high calibration accuracy.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of laser radar calibration, and specifically to a single-line and multi-line laser radar joint calibration method and system based on straight line features. Background Art

[0002] The purpose of joint calibration of single-line lidar and multi-line lidar is to calculate the relative posture between the single-line lidar and the multi-line lidar, that is, the yaw angle of the single-line lidar relative to the multi-line lidar and the translation in the x-axis and y-axis directions, so as to realize the transformation of the lidar coordinate system.

[0003] Among the currently available patents, Sany Heavy Industry Co., Ltd.'s "Calibration Method and Apparatus for Single-Line LiDAR and Multi-Line LiDAR" (Application Publication No. CN111103576 A) discloses a combined calibration method and apparatus for single-line and multi-line LiDARs. Its key feature is that the same point requires two coordinate transformations: first, the point in the single-line LiDAR coordinate system is transformed into the multi-line LiDAR coordinate system and then into the world coordinate system. This requires first calculating the pose of the multi-line LiDAR in the world coordinate system, and then calculating the relative pose between the single-line LiDAR and the multi-line LiDAR. The major drawback of this approach is that the two coordinate transformations introduce significant errors, ultimately leading to larger calibration errors. Furthermore, the approach is cumbersome to use. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: in response to the problems existing in the prior art, the present invention provides a single-line and multi-line laser radar joint calibration method and system based on straight line features that is easy to use and has high calibration accuracy.

[0005] In order to solve the above technical problems, the technical solution proposed by the present invention is:

[0006] A joint calibration method for single-line and multi-line lidar based on straight line features, comprising:

[0007] Perform preliminary cropping on the original multi-line lidar point cloud, fit two planes in the multi-line lidar point cloud, calculate the projection of the two planes in the xoy plane of the multi-line lidar coordinate system, thereby obtaining two straight lines, and then add several points along the z-axis direction;

[0008] Perform preliminary cropping on the original single-line lidar point cloud, fit two straight lines in the single-line lidar point cloud, and filter the single-line lidar point cloud so that the two straight lines in the single-line lidar point cloud are equal in length to the corresponding two straight lines in the multi-line lidar point cloud, and then add several points along the z-axis direction;

[0009] The single-line laser radar point cloud and the multi-line laser radar point cloud are aligned to obtain the relative position and posture of the single-line laser radar and the multi-line laser radar, and the calibration of the single-line laser radar and the multi-line laser radar is completed.

[0010] After obtaining the relative pose of the single-line laser radar and the multi-line laser radar, the accuracy of the relative pose is evaluated; when the accuracy of the relative pose is within the preset threshold, the calibration of the single-line laser radar and the multi-line laser radar is completed.

[0011] The accuracy evaluation process is as follows:

[0012] Get the normal vectors of the two lines in the multi-line lidar point cloud and the normal vectors of the two lines in the fitted single-line lidar point cloud, then calculate the angle between the two normal vectors corresponding to the same plane, get the corresponding rotation accuracy, and then get the average rotation error;

[0013] Obtain the intersection of two straight lines in the fitted single-line lidar point cloud, and the intersection of two straight lines in the multi-line lidar point cloud, and then calculate the distance between the two intersection points to obtain the translation accuracy;

[0014] Based on the average rotation error and translation accuracy, the calibration score is obtained;

[0015] The calibration score is compared with the preset standard calibration score. When the difference between the two is within the preset threshold, it means that the calibration result is valid.

[0016] The average rotation error Δψ is:

[0017]

[0018] Where Δψ1 and Δψ2 are the angular errors in two directions respectively;

[0019]

[0020] where n 2D and n' 2D They are the normal vectors of the two straight lines in the single-line lidar point cloud, n 3D and n' 3D are the normal vectors of the two lines in the multi-line lidar point cloud.

[0021] The translation accuracy Δt is:

[0022]

[0023] Among them, Δt is the distance error (i.e., translation accuracy), x1 and y1 are the coordinates of the intersection of two lines in the single-line lidar point cloud, and x2 and y2 are the coordinates of the intersection of two lines in the multi-line lidar point cloud.

[0024] As a further improvement of the above technical solution, the calibration score is:

[0025]

[0026] Wherein a and b are weighting coefficients, a+b=1.

[0027] The point cloud registration includes coarse registration and fine registration, and fine registration is performed after the coarse registration is completed. The specific process of obtaining the relative position of the single-line laser radar and the multi-line laser radar is as follows:

[0028] In the coarse registration stage, after the first coarse registration, the calculated transformation matrix is:

[0029]

[0030] After the second coarse registration, the calculated transformation matrix is:

[0031]

[0032] After the coarse registration is completed, the transformation matrix is ​​obtained:

[0033] T2=T1T0

[0034] In the fine registration stage, T2 is used as the initial value to perform the first fine registration. After the first fine registration is completed, the calculated transformation matrix is:

[0035]

[0036] Calculate the normal vectors of two lines in the single-line lidar and multi-line lidar point clouds using the singular value decomposition method. The calculated transformation matrix is:

[0037]

[0038] The final transformation matrix is:

[0039]

[0040] Among them, T0, T1, T2, T3, and T4 are all 4x4 augmentation matrices, R0, R1, R2, R3, and R4 are all 3x3 rotation matrices, and t0, t1, t2, t3, and t4 are all three-dimensional translation vectors;

[0041] At this point, the precise rotation matrix R and translation vector t are calculated, and the relative position of the single-line lidar relative to the multi-line lidar is obtained.

[0042] The process of calculating the projection of two planes in the xoy plane of the multi-line lidar coordinate system is:

[0043] Taking the Xoy plane of the LiDAR coordinate system as the reference, calculate the horizontal distance from each point in the preliminarily cropped multi-line LiDAR point cloud to the two straight lines. If the distance is less than the set threshold, retain the point. After traversing all point clouds, obtain a new point cloud. Set the vertical coordinate of each point in the new point cloud to 0. Extract the projection of the wall point cloud onto the Xoy plane of the multi-line LiDAR coordinate system, and then add several points along the z-axis.

[0044] The process of screening a single-line lidar point cloud is as follows: calculate the intersection of two straight lines in the fitted single-line lidar point cloud, calculate the distance from each point to the intersection, screen the point cloud based on the distance, and then add several points along the z-axis.

[0045] The present invention also discloses a single-line and multi-line laser radar joint calibration system based on straight line features, comprising:

[0046] The first program module is used to perform preliminary cropping on the original multi-line lidar point cloud, fit two planes in the multi-line lidar point cloud, calculate the projections of the two planes in the xoy plane of the multi-line lidar coordinate system to obtain two straight lines, and then add several points along the z-axis direction;

[0047] The second program module is used to perform preliminary cropping on the original single-line lidar point cloud, fit two straight lines in the single-line lidar point cloud, and filter the single-line lidar point cloud so that the two straight lines in the single-line lidar point cloud are equal in length to the corresponding two straight lines in the multi-line lidar point cloud, and then add several points along the z-axis direction;

[0048] The third program module is used to align the single-line laser radar point cloud with the multi-line laser radar point cloud, obtain the relative posture of the single-line laser radar and the multi-line laser radar, and complete the calibration of the single-line laser radar and the multi-line laser radar.

[0049] The present invention further discloses a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the calibration method described above are executed.

[0050] The present invention also discloses a computer device, comprising a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is run by the processor, the steps of the calibration method described above are executed.

[0051] Compared with the prior art, the advantages of the present invention are:

[0052] The present invention's method and system for joint calibration of single-line and multi-line laser radars based on straight-line features only requires one coordinate conversion, that is, converting from the single-line laser radar coordinate system to the multi-line laser radar coordinate system, directly performing point cloud registration, and obtaining the relative position and posture between the single-line laser radar and the multi-line laser radar, thereby achieving joint calibration of the single-line laser radar and the multi-line laser radar; since the above calibration method only performs one coordinate conversion, it is simple to operate compared to the multiple coordinate conversions in the existing methods, and can avoid calibration errors caused by multiple coordinate conversions, thereby improving calibration accuracy. In addition, the method described in the present invention does not require additional auxiliary tools and algorithms, and has low requirements for the scene in which data is collected. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a flowchart of an embodiment of the laser radar calibration method of the present invention.

[0054] Figure 2 This is a schematic diagram of the preliminary cropping of a single-line lidar point cloud according to the present invention.

[0055] Figure 3 This is a schematic diagram of the preliminary cropping of the multi-line laser radar point cloud of the present invention.

[0056] Figure 4 This is a schematic diagram (two planes) of the multi-line lidar point cloud after preliminary cropping of the present invention.

[0057] Figure 5 This is a schematic diagram of the point cloud preprocessing stage after adding new points along the z-axis at the intersection of the straight lines (the long lines in the figure represent the newly added points).

[0058] Figure 6 Schematic diagram of the accuracy evaluation of the present invention. DETAILED DESCRIPTION

[0059] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0060] like Figure 1 As shown, the single-line and multi-line laser radar joint calibration method based on straight line features according to an embodiment of the present invention includes:

[0061] Perform preliminary cropping on the original multi-line lidar point cloud, fit the two planes in the multi-line lidar point cloud, calculate the projection of the two planes in the xoy plane of the multi-line lidar coordinate system, and thus obtain two straight lines. Then add several points along the z-axis, such as Figure 5 As shown;

[0062] Perform preliminary cropping on the original single-line lidar point cloud, fit two straight lines in the single-line lidar point cloud, and filter the single-line lidar point cloud so that the two straight lines in the single-line lidar point cloud are equal in length to the corresponding two straight lines in the multi-line lidar point cloud, and then add several points along the z-axis direction;

[0063] Point cloud registration is performed on the single-line laser radar point cloud and the multi-line laser radar point cloud to obtain the relative position and posture of the single-line laser radar and the multi-line laser radar, and the calibration of the single-line laser radar and the multi-line laser radar is completed.

[0064] The single-line and multi-line laser radar joint calibration method based on straight line features of the present invention only needs to perform one coordinate conversion, that is, convert from the single-line laser radar coordinate system to the multi-line laser radar coordinate system, and directly perform point cloud alignment to obtain the relative posture between the single-line laser radar and the multi-line laser radar, thereby realizing the joint calibration of the single-line laser radar and the multi-line laser radar; since the above calibration method only performs one coordinate conversion, it is simple to operate compared with the multiple coordinate conversions in the existing method, and can avoid the calibration errors caused by multiple coordinate conversions, thereby improving the calibration accuracy.

[0065] In a specific embodiment, after obtaining the relative pose of the single-line laser radar and the multi-line laser radar, the accuracy of the relative pose is evaluated; if the accuracy of the relative pose is within a preset threshold, the calibration of the single-line laser radar and the multi-line laser radar is completed; otherwise, recalibration is required. The specific evaluation process is as follows:

[0066] Get the normal vectors of the two lines in the multi-line lidar point cloud and the normal vectors of the two lines in the fitted single-line lidar point cloud, then calculate the angle between the two normal vectors corresponding to the same plane, get the corresponding rotation accuracy, and then get the average rotation error;

[0067] Obtain the intersection of two straight lines in the fitted single-line lidar point cloud, and the intersection of two straight lines in the multi-line lidar point cloud, and then calculate the distance between the two intersection points to obtain the translation accuracy;

[0068] Based on the average rotation error and translation accuracy, the calibration score is obtained;

[0069] The calibration score is compared with the preset standard calibration score. When the difference between the two is within the preset threshold, it means that the calibration result is valid.

[0070] By evaluating the accuracy of the relative posture as described above, the reliability of subsequent calibration is guaranteed; and the above evaluation method is simple to operate and easy to implement.

[0071] In a specific embodiment, point cloud registration includes coarse registration and fine registration. Fine registration is performed after coarse registration is completed to finally obtain the relative position and posture of the single-line laser radar and the multi-line laser radar. The specific process is as follows:

[0072] In the coarse registration stage, after the first coarse registration, the calculated transformation matrix is:

[0073]

[0074] After the second coarse registration, the calculated transformation matrix is:

[0075]

[0076] After the coarse registration is completed, the transformation matrix is ​​obtained:

[0077] T2=T1T0

[0078] In the fine registration stage, T2 is used as the initial value to perform the first fine registration. After the first fine registration is completed, the calculated transformation matrix is:

[0079]

[0080] Calculate the normal vectors of two lines in the single-line lidar and multi-line lidar point clouds using the singular value decomposition (SVD) method. The calculated transformation matrix is:

[0081]

[0082] The final transformation matrix is:

[0083]

[0084] Among them, T0, T1, T2, T3, and T4 are all 4x4 augmentation matrices, R0, R1, R2, R3, and R4 are all 3x3 rotation matrices, and t0, t1, t2, t3, and t4 are all three-dimensional translation vectors;

[0085] At this point, the precise rotation matrix R and translation vector t are calculated, and the relative position of the single-line lidar relative to the multi-line lidar is obtained.

[0086] The above process has low requirements for the data collection scene, is simple and easy to operate, and can achieve accurate pose estimation, as shown below:

[0087] (1) When collecting data, only two intersecting planes are needed.

[0088] (2) During data preprocessing, the user only needs to draw two boxes and extract the point clouds of the two planes respectively. The remaining steps are all completed automatically, and the calibration results are directly output without any auxiliary tools or other auxiliary algorithms, such as SLAM (Simultaneous Localization And Mapping).

[0089] (3) During point cloud registration, the multi-line LiDAR point cloud and the single-line LiDAR point cloud are directly matched without the need for additional coordinate transformation. In addition, after two coarse registrations, a high-precision initial value can be obtained, and after two fine registrations, a high-precision calibration result, namely the rotation matrix and translation vector, can be obtained.

[0090] In a specific embodiment, the process of calculating the projection of two planes in the xoy plane of the multi-line lidar coordinate system is as follows: taking the xoy plane of the lidar coordinate system as a reference, calculating the horizontal distance from each point in the multi-line lidar point cloud after preliminary cropping to the two straight lines, if the distance is less than the set threshold, the point is retained; after traversing all point clouds, a new point cloud is obtained, the vertical coordinate of each point in the new point cloud is set to 0, and the projection of the wall point cloud to the xoy plane of the multi-line lidar coordinate system is extracted.

[0091] In a specific embodiment, the process of screening a single-line lidar point cloud is as follows: calculating the intersection of two straight lines in the fitted single-line lidar point cloud, calculating the distance from each point to the intersection, and screening the point cloud according to the size of the distance.

[0092] The embodiment of the present invention further discloses a single-line and multi-line laser radar joint calibration system based on straight line features, comprising:

[0093] The first program module is used to perform preliminary cropping on the original multi-line lidar point cloud, fit two planes in the multi-line lidar point cloud, calculate the projections of the two planes in the xoy plane of the multi-line lidar coordinate system to obtain two straight lines, and then add several points along the z-axis direction;

[0094] The second program module is used to perform preliminary cropping on the original single-line lidar point cloud, fit two straight lines in the single-line lidar point cloud, and filter the single-line lidar point cloud so that the two straight lines in the single-line lidar point cloud are equal in length to the corresponding two straight lines in the multi-line lidar point cloud, and then add several points along the z-axis direction;

[0095] The third program module is used to perform point cloud registration on the single-line laser radar point cloud and the multi-line laser radar point cloud, obtain the relative posture of the single-line laser radar and the multi-line laser radar, and complete the calibration of the single-line laser radar and the multi-line laser radar.

[0096] The combined calibration system of the present invention corresponds to the combined calibration method described above, and also has the advantages described in the calibration method described above.

[0097] like Figure 1 As shown, the above method is further described in detail below in conjunction with a complete specific embodiment:

[0098] Step 1: Point cloud preprocessing

[0099] like Figure 3 As shown in the figure, the multi-line laser radar point cloud is initially clipped, and only the point clouds within two planes (two intersecting and flat planes) are retained. The clipped point cloud is as follows Figure 4 shown.

[0100] The original single-line LiDAR point cloud is initially cropped to retain the two straight lines at the corner (such as Figure 2 The boxed part shown is included), and the rest is deleted;

[0101] Step 2: Extract straight lines

[0102] Fit two planes in the preliminarily cropped multi-line lidar point cloud, whose plane equations are:

[0103]

[0104] Among them, A1, B1, C1, D1 and A2, B2, C2, D2 are the parameters of the two plane equations. The two projection lines L1 and L2 of the two planes in the xoy plane of the multi-line lidar coordinate system are:

[0105]

[0106] Calculate the horizontal distance from each point in the initially cropped point cloud to lines L1 and L2 (based on the xoy plane of the multi-line LiDAR coordinate system). If the distance is less than the set threshold, retain the point. After traversing all points, a new point cloud is obtained. Set the vertical coordinate of each point in this new point cloud to 0. Then, extract the projection of the wall point cloud onto the xoy plane of the LiDAR coordinate system, and then add several points along the z-axis.

[0107] Then fit the two straight lines in the single-line lidar point cloud after preliminary cropping, calculate the intersection of the straight lines, and calculate the distance from each point to the intersection. According to the size of the distance, the point cloud is screened so that the two straight lines in the single-line lidar point cloud are equal in length to the two straight lines in the preprocessed multi-line lidar point cloud (that is, one straight line in the single-line lidar point cloud is equal in length to the straight line L1 in the multi-line lidar point cloud, and the other straight line in the single-line lidar point cloud is equal in length to the straight line L2 in the multi-line lidar point cloud). Then, several points are added along the z-axis to complete the preprocessing of the single-line lidar point cloud.

[0108] Step 3: Point cloud registration and pose estimation

[0109] Point cloud coarse registration calculates the initial value of the relative pose of the single-line laser radar and the multi-line laser radar; point cloud fine registration calculates the exact value of the relative pose of the single-line laser radar and the multi-line laser radar, and then calculates the normal vectors of the two straight lines in the point clouds of the single-line laser radar and the multi-line laser radar to optimize the accuracy of the relative pose.

[0110] Step 4: Accuracy Assessment

[0111] Specific description: Figure 6 As shown, the normal vectors of the two lines in the multi-line lidar point cloud are calculated, and the normal vectors of the two lines in the single-line lidar point cloud after fitting transformation are calculated. The angle between the two normal vectors corresponding to the same wall is calculated, and the rotation accuracy can be obtained:

[0112]

[0113] Among them, Δψ1 and Δψ2 are the angular errors in two directions (i.e., rotation accuracy), n 2D and n' 2D They are the normal vectors of the two straight lines in the single-line lidar point cloud, n 3D and n' 3D are the normal vectors of the two lines in the multi-line lidar point cloud.

[0114] The average rotation error is:

[0115]

[0116] Calculate the intersection of the straight lines in the converted single-line lidar point cloud, and calculate the intersection of the straight lines in the multi-line lidar point cloud. Calculate the distance between the two intersections to evaluate the translation accuracy, that is:

[0117]

[0118] Among them, Δt is the distance error (i.e., translation accuracy), x1 and y1 are the coordinates of the intersection of two lines in the single-line lidar point cloud, and x2 and y2 are the coordinates of the intersection of two lines in the multi-line lidar point cloud.

[0119] With the preset e r (rotation accuracy, unit: degree, for example 0.3 degrees) and e t (translation accuracy, unit: meter, for example 0.05 meter) as the benchmark, calculate the relative error, and calculate the weighted sum to get the calibration score:

[0120]

[0121] Where a and b are weighted coefficients, the sum of which is 1. Generally, a=0.7 and b=0.3 are selected according to the actual situation.

[0122] If the calibration score is less than 1, it means that the calibration result is valid.

[0123] This method is simple to use. Simply manually define the initial range (the point clouds on the two walls) in the original point cloud to directly output the calibration results, including the rotation matrix, translation vector, and calibration error of the single-line lidar relative to the multi-line lidar. Furthermore, this calibration method uses only a single coordinate transformation, resulting in high calibration accuracy. In a specific experiment, the average rotation accuracy of this calibration method was 0.228 degrees, the translation accuracy was 0.003 mm, and the calibration score was 0.533, which is significantly less than 1, indicating high calibration accuracy.

[0124] The embodiment of the present invention further discloses a computer-readable storage medium having a computer program stored thereon, which performs the steps of the calibration method described above when executed by a processor. The embodiment of the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, which performs the steps of the calibration method described above when executed by the processor. The present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. The memory can be used to store computer programs and / or modules. The processor implements various functions by running or executing the computer programs and / or modules stored in the memory, and accessing the data stored in the memory. The memory can include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0125] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A joint calibration method for single-line and multi-line lidar based on straight line features, which features include: Perform preliminary cropping on the original multi-line lidar point cloud, fit two intersecting planes in the multi-line lidar point cloud, calculate the projection of the two planes in the xoy plane of the multi-line lidar coordinate system, and thus obtain two straight lines. Then, add several points along the z-axis at the intersection of the two straight lines. Perform preliminary cropping on the original single-line LiDAR point cloud, fit two straight lines in the single-line LiDAR point cloud, and filter the single-line LiDAR point cloud so that the two straight lines in the single-line LiDAR point cloud are equal in length to the corresponding two straight lines in the multi-line LiDAR point cloud. Then, add several points along the z-axis at the intersection of the two straight lines in the single-line LiDAR point cloud; the two fitted straight lines in the single-line LiDAR point cloud correspond to the projection lines of the two planes of the multi-line LiDAR point cloud on the xoy plane. The single-line laser radar point cloud and the multi-line laser radar point cloud are aligned to obtain the relative position and posture of the single-line laser radar and the multi-line laser radar, and the calibration of the single-line laser radar and the multi-line laser radar is completed.

2. The single-line and multi-line laser radar joint calibration method based on straight line features according to claim 1 is characterized in that: After obtaining the relative pose of the single-line laser radar and the multi-line laser radar, the accuracy of the relative pose is evaluated; when the accuracy of the relative pose is within the preset threshold, the calibration of the single-line laser radar and the multi-line laser radar is completed.

3. The single-line and multi-line laser radar joint calibration method based on straight line features according to claim 2 is characterized in that: The evaluation process is as follows: Get the normal vectors of the two lines in the multi-line lidar point cloud and the normal vectors of the two lines in the fitted single-line lidar point cloud, then calculate the angle between the two normal vectors corresponding to the same plane, get the corresponding rotation accuracy, and then get the average rotation error; Obtain the intersection of two straight lines in the fitted single-line lidar point cloud, and the intersection of two straight lines in the multi-line lidar point cloud, and then calculate the distance between the two intersection points to obtain the translation accuracy; Based on the average rotation error and translation accuracy, the calibration score is calculated; The calibration score is compared with the preset standard calibration score. When the difference between the two is within the preset threshold, it means that the calibration result is valid.

4. The single-line and multi-line laser radar joint calibration method based on straight line features according to claim 3 is characterized in that: The average rotation error for: in and are the angular errors in two directions respectively; in and are the normal vectors of the two fitted straight lines in the single-line lidar point cloud, and are the normal vectors of the two fitted straight lines in the multi-line lidar point cloud.

5. The single-line and multi-line laser radar joint calibration method based on straight line features according to claim 4 is characterized in that: The translation accuracy for: in, is the distance error, and is the intersection coordinate of the two fitted straight lines in the single-line lidar point cloud, and is the coordinate of the intersection of two fitted lines in the multi-line lidar point cloud.

6. The single-line and multi-line laser radar joint calibration method based on straight line features according to claim 5 is characterized in that: The calibration score for: Where a and b are weighting coefficients, a+b=1.

7. The single-line and multi-line laser radar joint calibration method based on straight line features according to any one of claims 1 to 6, characterized in that: Point cloud registration includes coarse registration and fine registration. Fine registration is performed after coarse registration is completed. The specific process is as follows: In the coarse registration stage, after the first coarse registration, the calculated transformation matrix is: After the second coarse registration, the calculated transformation matrix is: After the coarse registration is completed, the transformation matrix is ​​obtained: In the fine registration stage, As the initial value, the first fine alignment is performed. After the first fine alignment is completed, the calculated transformation matrix is: Calculate the normal vectors of two lines in the single-line lidar and multi-line lidar point clouds using the singular value decomposition method. The calculated transformation matrix is: The final transformation matrix is: in, Both are 4x4 augmented matrices, Both are 3x3 rotation matrices, are all three-dimensional translation vectors; At this point, the precise rotation matrix R and translation vector t are calculated, and the relative position of the single-line lidar relative to the multi-line lidar is obtained.

8. The single-line and multi-line laser radar joint calibration method based on straight line features according to any one of claims 1 to 6, characterized in that: The process of calculating the projection of two planes in the xoy plane of the multi-line lidar coordinate system is: Taking the Xoy plane of the LiDAR coordinate system as the reference, calculate the horizontal distance from each point in the preliminarily cropped multi-line LiDAR point cloud to the two straight lines. If the distance is less than the set threshold, retain the point. After traversing all point clouds, obtain a new point cloud. Set the vertical coordinate of each point in the new point cloud to 0. Extract the projection of the wall point cloud onto the Xoy plane of the multi-line LiDAR coordinate system, and then add several points along the z-axis.

9. The single-line and multi-line laser radar joint calibration method based on straight line features according to any one of claims 1 to 6, characterized in that: The process of screening a single-line lidar point cloud is as follows: calculate the intersection of two straight lines in the fitted single-line lidar point cloud, calculate the distance from each point to the intersection, screen the point cloud based on the distance, and then add several points along the z-axis.

10. A single-line and multi-line laser radar joint calibration system based on straight line features, characterized in that: include: The first program module is used to perform preliminary cropping on the original multi-line lidar point cloud, fit two intersecting planes in the multi-line lidar point cloud, calculate the projections of the two planes in the xoy plane of the multi-line lidar coordinate system, thereby obtaining two straight lines, and then add several points along the z-axis at the intersection of the two straight lines; The second program module is used to perform preliminary cropping on the original single-line lidar point cloud, fit two straight lines in the single-line lidar point cloud, and filter the single-line lidar point cloud so that the two straight lines in the single-line lidar point cloud are of equal length to the corresponding two straight lines in the multi-line lidar point cloud. Then, a number of points are added along the z-axis at the intersection of the two straight lines in the single-line lidar point cloud; wherein the two fitted straight lines in the single-line lidar point cloud correspond to the projection lines of the two planes of the multi-line lidar point cloud on the xoy plane; The third program module is used to perform point cloud registration on the single-line laser radar point cloud and the multi-line laser radar point cloud, obtain the relative posture of the single-line laser radar and the multi-line laser radar, and complete the calibration of the single-line laser radar and the multi-line laser radar.

Images