Lidar inner parameter calibration method and inner parameter calibration system

Abstract

The application provides a laser radar inner parameter calibration method, comprising a calibration data acquisition step: acquiring multiple frames of point cloud data of different relative positions between a laser radar and a reflection plane of a target plate within a rated working range of the laser radar, and simultaneously, acquiring a spatial plane equation of the reflection plane 21 through a camera array corresponding to each frame of point cloud data; an inner parameter solving step: a target of inner parameter calibration in an inner parameter optimization model is to minimize the distance of each point on a calibration target plate measured by the laser radar to the plane of the target plate; and a mathematical optimization algorithm is used to solve an optimal solution of the above system under a constraint condition to determine calibration parameters of the laser radar. The laser radar inner parameter calibration method provides a global inner parameter integrated calibration method, obtains an inner parameter balancing correction accuracy under different scanning distances and scanning angles, and thus balances the performance of the laser radar within a rated scanning range, and avoids obvious performance accuracy reduction in an edge case.

Description

Technical Field

[0001] The present invention relates to the technical field of laser radar calibration, and in particular to a laser radar internal parameter calibration method and an internal parameter calibration system. Background Art

[0002] As an important component of modern sensor technology, LiDAR has been widely used in fields such as autonomous driving, robotics, and topographic mapping. LiDAR calculates the distance, speed, and direction of the target by emitting laser pulses and receiving the signals reflected by the object. However, due to installation, manufacturing, and structural constraints, there is a certain error between the actual measurement value and the theoretical value. In order to ensure the measurement accuracy and reliability of LiDAR, accurate internal parameter calibration of LiDAR is essential. Existing LiDAR calibration methods often rely on complex physical environment construction and tedious manual adjustments. This is not only time-consuming and labor-intensive, but also difficult to ensure the consistency and accuracy of the calibration results. Summary of the Invention

[0003] In order to improve the accuracy of laser radar internal parameter calibration and optimize the operation process of laser radar internal parameter calibration, the present invention provides a laser radar internal parameter calibration method and an internal parameter calibration system.

[0004] This application provides a laser radar internal parameter calibration method, including:

[0005] S1. Calibration data acquisition step: obtain multiple frames of point cloud data at different relative positions between the laser radar and the reflection plane of the target plate within the rated working range of the laser radar, where each frame of point cloud data is recorded as pc , the number of point cloud data frames is recorded as q At the same time, corresponding to each frame of point cloud data, the spatial plane equation of the reflection plane 21 is obtained through the camera array 3 n ; For each frame of point cloud data, calibration data is obtained d =( pc , n ). So the calibration data set can be expressed as ;

[0006] S2. Internal parameter calculation steps: The calibration parameters are α =( d ρ , d θ , d φ , s , h , v ),( d ρ , d θ , d φ ) is the distance, horizontal direction angle and vertical angle of the laser radar ( r , i , f ), s is the distance measurement scale factor, h and v To receive the horizontal and vertical offsets of the system from the origin of the device,

[0007] The correction equation of the measured value after calibration is:

[0008]

[0009] The goal of the internal parameter calibration in the internal parameter optimization model is to minimize the distance between each point on the calibration target plate measured by the lidar and the target plate plane. The objective function of the internal parameter optimization model is defined as:

[0010]

[0011] Where, P and Q k The calibration data obtained in the calibration data acquisition process, P is the set of all measured planes, Q k is the set of all points on the plane, d ( q , p k )for Q k midpoint q To plane p k Distance:

[0012]

[0013] In the above formula, x i 、 y i 、 z i for point q The coordinates of the measured values ​​are corrected by the calibration parameters α =( d ρ , d θ , d φ , s , h , v )Sure;

[0014] The optimal solution of the above system under the constraints is solved by mathematical optimization algorithm to determine the calibration parameters of the lidar α =( d ρ , d θ , d φ , s , h , v ).

[0015] Preferably, the S1 calibration data acquisition step is specifically as follows:

[0016] S11, adjust the laser radar inclination angle to i 1; Control the distance between the laser radar and the target plate to the set distance p 1; Collect laser radar i Scanned at 1 angle p 1 position target plate S Frame point cloud data, denoted as ;Calibration data in the current state , where n1 is the spatial plane equation parameter of the target plate in the current state.

[0017] S12, repeat S11 step, collect N The target plate scanning data of the laser radar within the preset distance interval between the target plate and the corresponding target plate space plane equation parameters are formed to form a calibration data set ;

[0018] S13, repeat steps S11-S12, collect M The target plate scanning data of the laser radar in the preset inclination range and the space plane equation parameters of the corresponding target plate are combined to form a complete calibration data set Q= .

[0019] Preferably, the parameters of the spatial plane equation of the target plate in the current state in the S1 calibration data acquisition step are obtained by solving the images obtained by imaging the reflection plane respectively by several cameras in the camera array at this time.

[0020] Preferably, the preset distance interval is 0.5-50m, and / or the preset inclination angle interval is -45-45°.

[0021] Preferably, the mathematical optimization algorithm in the S2 internal parameter solution step is the Levenberg-Marquardt algorithm.

[0022] Preferably, after the S2 internal parameter solution step is completed, the S3 internal parameter verification step is also performed: the relative position change between the target plate and the laser radar is controlled, and multiple frames of point cloud data corresponding to the reflection plane corrected using the calibration parameters are obtained. At the same time, the spatial plane equation of the reflection plane is obtained for each frame of point cloud data, and the average error of the point cloud data to the corresponding spatial plane equation is counted.

[0023] Preferably, if the average error is outside the threshold range, the calibration data acquisition step S1 and the internal parameter calculation step S2 are repeated to perform recalibration until the average error is within the threshold range.

[0024] This application also provides a laser radar internal parameter calibration system, including:

[0025] LiDAR, which has a fixed position;

[0026] The target plate has a varying spatial distance relative to the laser radar; a reflective surface with a finite area;

[0027] The camera array includes a plurality of cameras, and the cameras are used to combine and determine the spatial plane equation of the reflection plane.

[0028] Preferably, the laser radar can rotate around a fixed point in space, or the reflection plane can rotate around a fixed point in space.

[0029] Preferably, the target plate is movably arranged on a guide rail via a fixing frame; the guide rail extends in a direction away from the laser radar.

[0030] Alternatively, the target plate is placed on a fixed frame in a controlled rotational arrangement.

[0031] Alternatively, the target plate is movably arranged on a guide rail via a fixed frame; the guide rail extends in a direction away from the laser radar, and the cameras in the camera array are divided into two groups, and the two groups of cameras are placed on both sides of the guide rail, and the cameras in each group are arranged in sequence along the extension direction of the guide rail.

[0032] The laser radar internal parameter calibration method and internal parameter calibration system of the present application realize the automatic calibration of the laser radar. Through a pre-set process, the automatic unmanned calibration of the laser radar can be achieved. Compared with the existing laser radar calibration method, it can significantly improve the calibration efficiency and reduce labor input.

[0033] The lidar internal parameter calibration method of the present application provides a global internal parameter integration calibration method, which can take into account the different scanning conditions of the lidar and obtain internal parameters that balance the correction accuracy under different scanning distances and scanning inclination angles, thereby achieving a better balance in the performance of the lidar within the rated scanning range and avoiding a significant decrease in performance accuracy in edge cases. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1This is a schematic diagram of the architecture of the lidar intrinsic calibration system of this application;

[0035] Figure 2 This is a schematic diagram of the structure of the lidar internal reference calibration system of this application;

[0036] Figure 3 Schematic diagram of the drive of the target plate movement in the laser radar internal reference calibration system of this application;

[0037] Figure 4 Schematic diagram of the overall process of the lidar internal parameter calibration method of this application;

[0038] Figure 5 Schematic diagram of the automated calibration process of the lidar intrinsic calibration system of this application;

[0039] Figure 6 Schematic diagram of the reflective plane 21 of the target plate 2 in the laser radar intrinsic calibration system of the present application.

[0040] In the figure,

[0041] 1: LiDAR; 11: Calibration platform; 12: Pan / tilt platform; 2: Target plate; 21: Reflection plane; 22: Guide rail; 23: Fixing bracket; 3: Camera array; 31: Camera; 51: Integrated control unit; 52: Communication unit; 53: Computing and storage unit. DETAILED DESCRIPTION

[0042] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. In this specification, the size ratios in the drawings do not represent the actual size ratios, but are only used to reflect the relative positional relationship and connection relationship between the various components. Components with the same name or the same number represent similar or identical structures and are only for illustrative purposes.

[0043] This application mainly provides a highly automated lidar calibration system and a corresponding lidar intrinsic parameter calibration method. Figure 1 This is a schematic diagram of the architecture of the lidar calibration system. Figure 2 It is a structural diagram of the lidar calibration system.

[0044] LiDAR internal calibration system Figure 1As shown, the system comprises a fixed laser radar 1 and a movable target plate, as well as a camera array 3 for determining the precise spatial position of the target plate 2. The laser radar 1 is fixed in position, but optionally has a variable attitude, i.e., it can rotate in a fixed space. The target plate 2 provides a limited-area reflective surface 21 to reflect the laser beam from the laser radar 1. The target plate 2 is at least capable of translation in space to change its relative position to the target plate 2. The target plate 2 is preferably variable in attitude, i.e., it can rotate in a fixed space to change the orientation of the reflective surface 21. The camera array 3 includes several strategically arranged cameras 31. Based on the combined images of the cameras 31, the position and orientation of the reflective surface 21 are determined. This is the plane equation of the reflective surface 21 in the spatial coordinate system. To ensure the precise determination of the specific position of the reflective surface 21 within the image captured by the cameras 31, a plurality of checkerboard patterns are arranged on the surface of the reflective surface 21. The checkerboard patterns are preferably located at the four corners of the reflective surface 21. The target plate 2 is a rectangular flat plate with a flat surface and coated with a high-laser reflectivity coating, providing a stable reflection signal for laser radar calibration. The four corners of target plate 2 are labeled with camera array tags, providing feature points for the camera array. It can be understood that the movement of target plate 2 is equivalent to the posture change of lidar 1 to a certain extent. In other words, if the orientation of target plate 2 is ignored, the posture change of lidar 1 can be equivalent to the displacement of target plate 2 in the same direction.

[0045] The purpose of the camera array 3 including multiple cameras 31 is to calculate the spatial position of the reflection plane 21 through planar imaging of the multiple cameras 31 and to obtain the spatial plane equation of the reflection plane 21. Another purpose is to ensure that the camera 31 is always within a reasonable distance from the target plate 2 within the range of motion of the target plate 2 through reasonable layout, thereby improving the positioning accuracy of the target plate 2 within the range of motion.

[0046] Figure 2 A schematic diagram of the structure of a laser radar calibration system that meets the above conditions is given. A calibration platform 11 is set in the calibration space, and the laser radar 1 is set on the calibration platform 11 for fixed movement. The calibration platform 11 is a device for supporting and adjusting the position and posture of the laser imaging radar equipment. The laser radar 1 is set on the calibration platform 11 through a pan-tilt platform 12. The pan-tilt platform 12 provides multiple degrees of freedom of movement, which can enable the laser radar to be accurately leveled in multiple degrees of freedom to meet the requirements of the laser radar internal parameter calibration. Preferably, the pitch and left and right swing of the laser radar 1 are guaranteed.

[0047] The target plate 2 is movably mounted on a guide rail 22 via a mounting bracket 23. The guide rail 22 extends away from the laser radar 1. When the mounting bracket 23 is controlled to move on the guide rail 22, the relative distance between the reflective surface 21 and the laser radar 1 changes. The target plate 2 is preferably mounted on the mounting bracket 23 in a controlled rotational manner, meaning that the specific orientation of the reflective surface 21 on the mounting bracket 23 can be controlled and adjusted. Figure 3Figure 2 is a schematic diagram of the motion mechanism of target plate 2. A rack-and-pinion drive driven by a stepper motor enables movement of the fixed frame 23 on the guide rail 22, while a worm gear mechanism driven by the stepper motor enables the target plate 2 to change its position on the fixed frame 23. Target plate 2 provides a stable target for LiDAR intrinsic calibration, ensuring its accuracy and reliability.

[0048] The distance between target plate 2 and laser radar 1 can be adjusted by adjusting the distance of target plate 2, and the angle of the reflection surface in different scenarios can be simulated by changing the orientation of target plate 2. At the same time, the variable posture of laser radar 1 ensures that any scanner on laser radar 1 can always obtain its effective reflection point on reflection plane 21 through combined control to achieve the purpose of posture correction.

[0049] The cameras 31 are divided into two groups, one on each side of the guide rail 22. The cameras 31 in each group are arranged sequentially along the extension direction of the guide rail 22. The diagram shows a situation with three cameras per group. At any position on the target plate 2, at least two opposing cameras 31 can image the reflection plane 21, thereby solving the spatial plane equation of the reflection plane 21. In most cases, four to six groups of cameras 31 can image the reflection plane 21 to improve the accuracy of the displacement solution for the reflection plane 21. Extending along the guide rail 22 primarily ensures that there is a camera 31 relatively close to the target plate 2 throughout the entire range of motion of the target plate 2 along the guide rail 22, thereby ensuring accuracy.

[0050] Camera array 3 comprises six cameras, fixedly mounted in the experimental environment and distributed on either side of the motion track. By identifying and measuring visual positioning tags affixed to the target plate, they achieve real-time, precise measurement of the plane pose (i.e., plane equation parameters) of the target plate 2. Multiple visual markers are placed on the target plate 2, each with recognizable geometric features to facilitate capture by the camera array. Each camera undergoes intrinsic and extrinsic calibration to obtain its internal parameters and its position and pose information relative to the global coordinate system. Each camera captures an image of the visual marker on the target plate 2, and the visual markers in each image are detected and matched. Based on the matching results of the visual markers in each camera image, the three-dimensional coordinates of the visual markers in the global coordinate system are calculated. Using the three-dimensional coordinate information of the visual markers, the least squares method is used to calculate the equation of the spatial plane in which the target plate 2 resides. The spatial plane equation is expressed as ax+by+cz+d=0, where a, b, and c are the normal vector components of the plane, and d represents the distance from the plane to the origin. The parameters a, b, c, and d of the target plate 2 plane equation measured by the camera array 3 provide high-precision visual auxiliary information for the internal calibration of the laser radar 1.

[0051] The computational storage unit 53 is a computer integrating a high-performance processor and storage unit. It is responsible for calculating, analyzing, and storing the received LiDAR data, while also running the calibration algorithm to calculate the LiDAR's internal parameters. The integrated control unit 51 coordinates and controls the entire calibration system. The communication unit 52 provides a data transmission channel, including motion control of the fixed frame 23, motion control of the target plate 2, data reading from the camera array 3, and real-time transmission of LiDAR 1 data.

[0052] The laser radar internal parameter calibration algorithm of this application is applied to the above laser radar internal parameter calibration system. Figure 1 As shown, the steps include:

[0053] S1. Calibration data acquisition step. Within the rated operating range of the laser radar, multiple frames of point cloud data at different relative positions between the laser radar 1 and the reflection plane 21 are obtained, where each frame of point cloud data is recorded as pc and the number of point cloud data frames is recorded as q. At the same time, corresponding to each frame of point cloud data, the spatial plane equation n of the reflection plane 21 is obtained through the camera array 3. For each frame of point cloud data, the calibration data d=(pc,n) is obtained. The calibration data set can then be expressed as .

[0054] Specifically,

[0055] S11, install the laser radar to be calibrated on the calibration platform 11 through the pan-tilt platform, and adjust the angle of the pan-tilt platform so that the laser radar to be calibrated and the calibration platform form a certain angle i 1; Through the integrated control unit and the communication unit, the control frame 23 carries the target plate 2 and moves to a specific position along the guide rail 22 p 1; Through the integrated control unit 51 and the communication unit 52, the laser radar 1 to be calibrated is collected i Scanned at 1 angle p 1 position target plate 2 S Frame point cloud data, denoted as Through the integrated control unit 51 and the communication unit 52, the camera array 3 is controlled to solve the plane equation parameters of the target plate 2 relative to the global coordinate system in the current state ; The calibration data in the current state is stored by the calculation storage unit (210) .

[0056] S12, repeat the above S11 step, collect N The target plate 2 scanning data and corresponding plane equation parameters of the laser radar 1 to be calibrated within the preset distance interval between the laser radar 1 and the target plate 2 are combined to form a calibration data set. ;

[0057] S13, repeat steps S11-S12, collectM The scanning data of the target plate 2 of the laser radar 1 to be calibrated within the preset inclination range and the corresponding plane equation parameters are combined to form a complete calibration data set ;

[0058] For general lidar, the preset distance range is 0.5-50m, and the preset inclination angle range is -45-45°.

[0059] S2. Internal parameter calculation step.

[0060] The internal calibration model includes 6 parameters: distance to the laser radar, horizontal angle and vertical angle ( r , i , f ) error is calibrated, and the calibration parameter is the corresponding offset value α = ( d ρ , d θ , d φ ), and the distance measurement scale factor s , receiving the horizontal and vertical offset of the system from the device origin h and v Therefore, the calibration parameters are α =( d ρ , d θ , d φ , s , h , v ), the correction equation of the measured value after calibration is:

[0061]

[0062] For laser radar, each scanner has its own calibration parameters. Therefore, for a 64-line rotating laser radar, the number of its internal parameters is 6×64; for micro-electromechanical system solid-state radar, its parameters are similar to those of mechanical type; for array solid-state radar, such as m × n pixels, and the number of its intrinsic parameters is m × n ×6.

[0063] Constructing an internal parameter optimization model: The goal of internal parameter calibration is to minimize the distance between each point on the calibration target plate measured by the LiDAR and the target plate plane. In other words, when the LiDAR scans a plane, the degree of discreteness of the point cloud on the plane should be minimized. The objective function of the internal parameter optimization model can be defined as:

[0064] (2)

[0065] In the above formula, P and Q k The calibration data obtained in the calibration data acquisition process, P is the set of all measured planes, Q k is the set of all points on the plane, d ( q , p k )for Q k midpoint q To plane p k Distance:

[0066]

[0067] In the above formula, x i 、 y i 、 z i for point q The coordinates of the measured values ​​can be expressed in terms of calibration parameters by applying the correction equations for the measured values. α =( d ρ , d θ , d φ , s , h , v )express.

[0068] The internal parameter solution, whose minimum distance is an optimization problem, can be solved using existing mathematical optimization methods to obtain the optimal solution of the system under the above constraints, thereby determining the optimal values ​​of the calibration parameters. For example, the Levenberg–Marquardt algorithm can be used, using the design parameters as initial values, and then iteratively solving to obtain the calibration values ​​of the calibration parameters that meet the constraints.

[0069] Because the above method treats the multiple frames of point cloud data as a system for an iterative solution, it helps achieve a global optimum during the LiDAR point cloud calibration process, avoiding the local optimum issues caused by a divide-and-conquer approach. Conventional methods use the least squares method to calculate the correction value of each point in the point cloud, thereby determining the corresponding internal parameters of the scanner.

[0070] The method of this application combines a collection of point cloud data from different frames under various preset scenarios within the rated operating range of the lidar 1 to perform a holistic solution to achieve a global optimal solution. Obviously, for the same scanner, its correction value may vary depending on the distance and angle between the lidar 1 and the reflective plane 21. In other words, the correction value of the scanned data corresponding to the same scanner in different states may vary slightly. Because the solution of this application is a holistic solution and implicitly utilizes constraints, it can simultaneously optimize the global parameters of the scanner to achieve the goal of obtaining a global optimal solution.

[0071] Optionally, S3, an internal reference verification step is also included.

[0072] Through the integrated control unit 51 and the communication unit 52, the fixed frame carrying the target plate 2 is controlled to move along the guide rail 22 to other positions different from the position when the data is acquired; through the integrated control unit 51 and the communication unit 52, the point cloud data of the target plate 2 at the position obtained by scanning the laser radar 1 to be calibrated is collected; through the integrated control unit 51 and the communication unit 52, the camera array 3 is controlled to solve the plane equation parameters of the target plate 2 relative to the global coordinate system in the current state; the distance from all target plate measurement points to the target plate plane is calculated, and the average distance error is counted; repeat multiple times to determine whether the distance errors of all positions are within an acceptable range, that is, whether they are within a preset threshold. If acceptable, stop calibration; if unacceptable, add calibration data or re-collect calibration data to recalibrate the internal parameters of the laser radar 1 until the requirements are met.

[0073] The automatic calibration process of the above-mentioned laser radar internal parameter automatic calibration system is as follows: Figure 5 .

[0074] 1. System Preparation: First, ensure that the target plate and motion rails, calibration target plate, camera array, calibration platform and gimbal, computing and storage unit, communication unit, and integrated control unit are all installed and debugged. Check all hardware devices to ensure that the calibration target plate and lidar are initially positioned correctly.

[0075] 2. System startup: Start the integrated control unit and enter the calibration task process. Start the communication unit to ensure smooth communication between systems.

[0076] 3. Calibration task setting: Set the specific parameters of the calibration task in the integrated control unit, such as calibration task type, calibration accuracy requirements, etc.

[0077] 4. Calibration target plate identification and data acquisition: Start the target plate motion system and move the calibration target plate according to the preset task so that it passes through the scanning area of ​​the laser radar. When the target plate stops at each calibration task point in the scanning area of ​​the laser radar 1, the target plate accurately rotates the calibration target plate to the angle required by the instruction according to the instruction requirements of the integrated control unit. The laser radar 1 scans the calibration target plate, collects laser point cloud data, and transmits it to the computing and storage subsystem in real time. The camera arrays distributed on both sides of the motion guide rail perform real-time and accurate measurement of the target plate's posture by identifying and measuring the visual positioning tags fixed on the target plate. The target plate plane equation parameters obtained by the camera array processing are transmitted to the computing and storage unit through the communication unit for solving the laser radar's calibration internal parameters.

[0078] 5. Data Processing and Parameter Optimization: After receiving the data, the computational storage unit processes it according to a preset algorithm, identifies target features, and calculates the lidar's internal calibration parameters. Multi-parameter, high-precision optimization is performed on the collected data to ensure the accuracy of the calibration results.

[0079] 6. Calibration Verification: Use the calibration target to verify the LiDAR calibration results and check their accuracy and reliability. Adjust the calibration parameters according to the accuracy setting requirements and repeat the above steps until the LiDAR achieves satisfactory calibration accuracy.

[0080] 7. Calibration result storage and report generation: Store the final calibration parameters in the calculation storage unit and generate a detailed calibration report.

[0081] 8. System shutdown and maintenance: After completing the calibration task, shut down all subsystems and perform necessary maintenance work, such as cleaning optical components and checking mechanical parts, to ensure that the system can operate normally the next time it is used.

[0082] The above content merely describes preferred embodiments of the present invention and does not limit the scope of the present invention. Any modifications and improvements to the technical solution of the present invention made by persons skilled in the art without departing from the spirit of the present invention shall fall within the scope of protection defined by the claims.

Claims

1. A laser radar internal parameter calibration method, characterized in that: include: S1. Calibration data acquisition step: obtain multiple frames of point cloud data at different relative positions between the laser radar and the reflection plane of the target plate within the rated working range of the laser radar, where each frame of point cloud data is recorded as pc , the number of point cloud data frames is recorded as q At the same time, corresponding to each frame of point cloud data, the spatial plane equation of the reflection plane 21 is obtained through the camera array 3 n ; For each frame of point cloud data, calibration data is obtained d =( pc , n ); the calibration data set can be expressed as ; S2. Internal parameter calculation steps: The calibration parameters are α =( δ ρ , δ θ , δ φ , s , h , v ),( δ ρ , δ θ , δ φ ) is the distance, horizontal direction angle and vertical angle of the laser radar ( ρ , θ , φ ), s is the distance measurement scale factor, h and v To receive the horizontal and vertical offsets of the system from the origin of the device, The correction equation of the measured value after calibration is: The goal of the internal parameter calibration in the internal parameter optimization model is to minimize the distance from each point on the calibration target plate measured by the lidar to the target plate plane. The objective function of the internal parameter optimization model is defined as: Where, P and Q k The calibration data obtained in the calibration data acquisition process, P is the set of all measured planes, Q k is the set of all points on the plane, d ( q , p k )for Q k midpoint q To plane p k Distance: In the above formula, x i 、 y i 、 z i for point q The coordinates of the measured values ​​are corrected by the calibration parameters α =( δ ρ , δ θ , δ φ , s , h , v )Sure; The optimal solution of the above system under the constraints is solved by mathematical optimization algorithm to determine the calibration parameters of the lidar α =( δ ρ , δ θ , δ φ , s , h , v ).

2. The laser radar internal parameter calibration method according to claim 1, wherein: The S1 calibration data acquisition steps are as follows: S11, adjust the laser radar inclination angle to θ 1; Control the distance between the laser radar and the target plate to the set distance p 1; Collect laser radar θ Scanned at 1 angle p 1 position target plate S Frame point cloud data, denoted as ;Calibration data in the current state , where n1 is the spatial plane equation parameter of the target plate in the current state; S12, repeat S11 step, collect N The target plate scanning data of the laser radar within the preset distance interval between the target plate and the corresponding target plate space plane equation parameters are formed to form a calibration data set ; S13, repeat steps S11-S12, collect M The target plate scanning data of the laser radar in the preset inclination range and the space plane equation parameters of the corresponding target plate are combined to form a complete calibration data set Q= .

3. The laser radar internal parameter calibration method according to claim 2, characterized in that: The parameters of the spatial plane equation of the target plate in the current state in the S1 calibration data acquisition step are obtained by solving the images obtained by imaging the reflection plane respectively by several cameras in the camera array at this time.

4. The laser radar internal parameter calibration method according to claim 2, wherein: The preset distance interval is 0.5-50m, and / or the preset inclination angle interval is -45-45°.

5. The laser radar internal parameter calibration method according to claim 2, wherein: The mathematical optimization algorithm in the S2 internal parameter solution step is the Levenberg-Marquardt algorithm.

6. The laser radar internal parameter calibration method according to claim 1, characterized in that: After the S2 internal parameter solution step is completed, the S3 internal parameter verification step is also executed: the relative position change between the target plate and the laser radar is controlled, and multiple frames of point cloud data corresponding to the reflection plane corrected by the calibration parameters are obtained. At the same time, the spatial plane equation of the reflection plane is obtained for each frame of point cloud data, and the average error of the point cloud data to the corresponding spatial plane equation is counted.

7. The laser radar internal parameter calibration method according to claim 6, wherein: If the average error is outside the threshold range, repeat the calibration data acquisition step S1 and the internal parameter solution step S2 to recalibrate until the average error is within the threshold range.

8. A laser radar internal parameter calibration system using a laser radar internal parameter calibration method according to any one of claims 1 to 7, characterized in that: include Laser radar (1), which has a fixed position; A target plate (2) having a varying spatial distance relative to the laser radar 1; and a reflecting plane (21) having a limited area; The camera array (3) includes a plurality of cameras (31), and the cameras (31) are used to combine and determine the spatial plane equation of the reflection plane (21).

9. The laser radar internal parameter calibration system according to claim 8, characterized in that: The laser radar (1) can rotate around a fixed point in space, or the reflection plane (21) can rotate around a fixed point in space.

10. The laser radar internal parameter calibration system according to claim 8, characterized in that: The target plate (2) is movably arranged on the guide rail (22) via the fixing frame (23); the guide rail (22) extends in a direction away from the laser radar (1). Alternatively, the target plate (2) is arranged on a fixed frame (23) in a controlled rotational manner. Alternatively, the target plate (2) is movably arranged on the guide rail (22) via a fixing frame (23); the guide rail (22) extends in a direction away from the laser radar (1); the cameras (31) in the camera array (3) are divided into two groups, and the two groups of cameras (31) are respectively placed on both sides of the guide rail (22), and the cameras (31) in each group are sequentially spaced along the extension direction of the guide rail (22).

Images