Swing scanning type laser panoramic three-dimensional scanning device based on high-precision rotary table and panoramic measurement method

Abstract

The invention discloses a swing scanning type laser panoramic three-dimensional scanning device based on a high-precision rotary table and a panoramic measurement method, the swing scanning type laser panoramic three-dimensional scanning device based on the rotary table is used for realizing laser scanning, and a camera only collects line contour information of one surface area every time; the problem of plane mirror light path crosstalk formed in different ROI regions is effectively avoided. Line laser can be projected to the plane mirror at any angle, slight front-and-back pitching of the plane mirror does not affect the measurement result, the installation requirement is low, and the device has the advantages of higher spatial resolution, high precision, large scanning range, high measurement scene applicability, high measurement efficiency and wide application range. Compared with a conventional linear scanning mode, the swing scanning type laser scanning device has the advantage of higher information acquisition speed, and when a sample with a large size is reconstructed under a large working distance, the information acquisition efficiency based on the swing scanning type laser scanning mode is far higher than that of translation scanning.

Description

Technical Field

[0001] The present invention relates to the field of optical measurement, and in particular to a panorama three-dimensional scanning device and a panorama measurement method based on a high-precision turntable. Background Art

[0002] As a practical and powerful non-contact shape measurement method, active optical 3D measurement technology has been widely used in cultural relics protection, reverse engineering, industrial design, inspection and other fields. Based on the high brightness and high collimation of lasers, active optical 3D measurement devices based on laser scanning have achieved rapid development.

[0003] Traditional panoramic three-dimensional measurement technology often uses multiple sets of cameras to collect three-dimensional information of the sample surface from different angles, or moves the sample to collect three-dimensional information of the sample from different viewing angles. In practical applications, it is costly, complex to set up, and cumbersome to calibrate. In order to reduce costs and simplify settings, a good solution is to use a mirror-assisted multi-view measurement system to achieve panoramic three-dimensional profile measurement. This method is equivalent to arranging multiple sets of measuring devices (i.e., a set of actual measuring devices and multiple virtual measuring devices imaged by plane mirrors) in a surrounding configuration, and collecting object surface data from multiple viewing angles at the same time. This solution only requires one set of measuring devices and several plane mirrors to achieve 360-degree panoramic three-dimensional surface measurement. Many scholars have conducted in-depth research on the projection method of mirror-assisted panoramic three-dimensional measurement systems. Some scholars have constructed an orthogonal projection system using Fresnel lenses and commercial DLP projectors to generate plane-parallel projection light, and some scholars have introduced a fringe projection profiler system into a plane mirror to achieve high-speed panoramic three-dimensional shape measurement.

[0004] However, the existing mirror-assisted panoramic 3D measurement technology still faces some problems:

[0005] 1. Due to the introduction of the plane mirror, the sample to be measured is imaged in three different areas. The optical 3D measurement technology based on active light source will project stripes on the surface of the three areas. The depth information of the three areas has certain differences, and the plane mirror imaging will cause light crosstalk under different viewing angles. The quality of the stripe pattern collected by the camera has a significant impact on the accuracy of the measurement system.

[0006] 2. The object to be measured is imaged in three areas through a plane mirror. The camera needs to collect three-dimensional information of multiple surface areas at the same time. The camera's pixels are divided into multiple image areas, and the effective resolution is greatly reduced.

[0007] 3. The mirror-assisted panoramic scanning system based on surface structured light projection and flat-scan laser scanning is mainly suitable for the three-dimensional reconstruction of small-sized objects. For large-sized objects to be measured, the existing technology is difficult to achieve compatibility between high-precision and high-efficiency measurement. Summary of the invention

[0008] In response to the need for improvement in the prior art, the present invention provides a panoramic three-dimensional laser scanning device and a panoramic measurement method based on a high-precision turntable. By introducing a plane mirror, an object can be imaged at three viewing angles, thereby achieving 360-degree panoramic three-dimensional measurement of the surface of the object to be measured. At the same time, the swing-scanning laser measurement system based on the turntable can sequentially scan three different areas of the sample surface, thereby obtaining three-dimensional information of the panoramic surface of the sample in a single scanning sequence. The scanning method using a rotating laser beam enhances the system's applicability to samples of different scales and improves the efficiency of three-dimensional data acquisition, making the system more suitable for surface reconstruction under a large field of view.

[0009] To achieve the above objectives, the technical methods adopted by the present invention are specifically as follows:

[0010] The present invention discloses a swing-sweep laser panoramic three-dimensional scanning device based on a high-precision turntable, comprising a camera, a line laser, a turntable, a controller, a computer for data processing connecting the controller and the camera, and two plane mirrors for collecting multi-view images. The line laser and the camera are fixed on a rotating bracket on the turntable. There is a horizontal distance difference between the line laser and the camera to generate the parallax required for three-dimensional reconstruction. The two plane mirrors for collecting multi-view images are placed behind the object to be measured and form an angle. The turntable is connected to the controller.

[0011] As a further improvement, the line laser described in the present invention projects a laser surface onto a plane mirror at an arbitrary angle.

[0012] As a further improvement, the angle between the two plane mirrors described in the present invention is between 90 degrees and 180 degrees.

[0013] As a further improvement, the angle described in the present invention is 120 degrees.

[0014] The present invention also discloses a panoramic measurement method of a swing-sweep laser panoramic three-dimensional scanning device based on a high-precision turntable, comprising:

[0015] Place the calibration plate in N positions, collect N pictures of the calibration plate through the camera, turn on the laser, and take N pictures of the laser stripe pattern projected onto the corresponding calibration plate;

[0016] Extract the coordinates of the corner points in the collected calibration images, and use Zhang's calibration method to calculate the internal and external parameters of the camera to obtain the plane equation of the calibration plate under different postures;

[0017] The Steger centerline extraction algorithm is used to extract the pixel coordinates on the laser stripe image. The three-dimensional coordinates of the laser line are obtained by combining the pixel coordinates of the laser line and the plane equation of the corresponding calibration plate. Finally, the three-dimensional coordinates of the laser line under N different postures are used to fit the laser plane equation.

[0018] Fix the camera on the rotating table where the laser is located, place the calibration plate in N positions, use the camera to simultaneously shoot the calibration plate and its virtual image in the mirror, extract the three-dimensional feature point pairs of the calibration plate and the virtual calibration plate image in the mirror, and then use the Levenberg-Marquardt algorithm with bundle adjustment strategy to calculate the plane equation of the plane mirror, so as to obtain the reflection matrix of the plane mirror in the global camera coordinate system at the initial position;

[0019] Turn on the rotating stage. Every time the stage rotates a certain angle around the rotation axis, the camera collects a calibration plate picture. After obtaining N calibration plate pictures, the PnP algorithm is used in combination with the camera's intrinsic parameters to obtain the extrinsic parameters of the chessboard in each picture, and the three-dimensional coordinates of the chessboard corner points in their corresponding camera coordinate system are calculated. The initial parameters of the rotating stage are calculated using the three-dimensional coordinates of the chessboard corner points at different camera positions. Finally, the rotation transformation matrix of the rotating stage in the global camera coordinate system at the initial position is obtained through optimization using the Levenberg-Marquardt algorithm.

[0020] Using the plane mirror reflection matrix at the initial position and the rotation transformation matrix of the turntable, calculate the reflection matrix of the plane mirror in the local camera coordinate system when the turntable moves to different angles;

[0021] Turn on the rotating stage and the laser, and use the camera to sequentially collect the laser stripe patterns of the sample in the left plane mirror imaging area, the front surface area, and the right plane mirror imaging area during scanning;

[0022] The laser stripe patterns of the left plane mirror imaging area, the front surface area and the right plane mirror imaging area of ​​the sample are converted into three-dimensional line profile information of the left plane mirror area, the front surface area and the right plane mirror area using the obtained camera intrinsic parameters and the calibrated laser plane equation;

[0023] The three-dimensional line profile information of the virtual surfaces of the left plane mirror area and the right plane mirror area is transformed into the real point cloud coordinates in the local camera coordinate system of the corresponding angle through the reflection matrix of the plane mirror;

[0024] Use the rotation transformation matrix of the rotating stage to transform the real point cloud coordinates of the three areas in the local camera coordinate system at different angles into the point cloud coordinates in the global camera coordinate system at the initial position;

[0025] The point cloud coordinates obtained in the initial camera coordinate system in different areas are merged to obtain the panoramic three-dimensional contour information of the sample surface.

[0026] As a further improvement, the rotation transformation matrix of the rotation stage in the initial position global camera coordinate system of the present invention is specifically:

[0027] The relationship between the three-dimensional coordinates of a corner point of a calibration plate in the camera coordinate system of two adjacent cameras can be expressed as:

[0028]

[0029] Among them, T is the rotation transformation matrix between the two coordinate systems, expressed as:

[0030]

[0031] After obtaining N groups of rotations around the rotation axis with a rotation interval of θ c After obtaining the three-dimensional point coordinates, let G = {uvwabc} and construct the equation:

[0032] G=argmin(g1(G)+g2(G)+·······+g N (G)

[0033]

[0034] Minimizing the above formula is a nonlinear minimization problem. The parameters of the rotation axis can be calculated by iteratively solving the Levenberg-Marquardt algorithm.

[0035] As a further improvement, the reflection matrix of the plane mirror in the local camera coordinate system when the rotating stage of the present invention moves to different angles is specifically:

[0036] The camera coordinate system (X0, Y0, Z0) at the initial position of the rotating stage is used as the global coordinate of the system, and the camera coordinate system (X i ,Y i ,Z i ) as the i-th local coordinate system, the two coordinate systems

[0037] The relationship between can be expressed as:

[0038]

[0039] Among them, θ i The camera is at position P i The angle of rotation from the reference position P0,

[0040] Since the relative positions of the laser plane and the camera are fixed, the light plane equation coefficients of the laser plane do not change when the camera position changes;

[0041] When the camera is at the reference position P0, the mirror equation on one side is described as:

[0042]

[0043] According to the relationship between the two coordinate systems, when the camera is at position P i When , the mirror equation of the plane mirror can be described as:

[0044] [a i b i c i d i ]=[a 0 b 0 c 0 d 0 ]T(θ i ) -1

[0045] Then, in the i-th local coordinate system, the reflection matrix that transforms the point cloud coordinates of the virtual surface of the plane mirror to the real position can be expressed as:

[0046]

[0047] in

[0048] As a further improvement, the rotation transformation matrix of the rotating stage described in the present invention converts the real point cloud coordinates of the three regions in the local camera coordinate system at different angles into the point cloud coordinates in the global camera coordinate system at the initial position. Specifically, the camera coordinate system (X0, Y0, Z0) at the initial position of the rotating stage is used as the global coordinate of the system, and the camera coordinate system (X0, Y0, Z0) at the i-th scanning position is used as the global coordinate of the system. i ,Y i ,Z i ) is taken as the i-th local coordinate system, and the relationship between the two coordinate systems can be expressed as:

[0049]

[0050] Among them, θ i The camera is at position P i The angle of rotation from the reference position P0;

[0051] According to the relationship between the two coordinate systems, the relationship between the local coordinates in the i-th local coordinate system and the global coordinates in the global coordinate system is:

[0052]

[0053] Among them, θ i The camera is located at the local coordinate position P iThe angle rotated from the reference position P0 to the reference position P0.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] 1. Simple installation and low cost: The present invention proposes a turntable-based swing-sweep method to achieve laser scanning. The camera only collects line profile information of one surface area each time, effectively avoiding the problem of plane mirror optical path crosstalk formed in different ROI areas. Line laser can be projected onto the plane mirror at any angle, and the slight pitch of the plane mirror will not affect the measurement results, and the installation requirements are relatively low.

[0056] 2. Higher spatial resolution and high precision: Conventional 3D panoramic reconstruction methods require each camera to capture 3D information of the surface from multiple perspectives at the same time. The camera's pixels are divided into multiple image regions, and the effective resolution of the reconstruction result is significantly reduced. However, in this method, the camera captures the 3D information of three surfaces sequentially, and each image contains only the line profile information of one surface. Therefore, the system can achieve higher resolution under the same camera parameters.

[0057] 3. Large scanning range and strong applicability in measurement scenarios: The size of the swing-sweeping laser scanning device proposed in the present invention does not need to match the size of the scanned sample. Only a small-sized rotating table is needed to achieve three-dimensional reconstruction of samples of different sizes, thus avoiding the influence of the sample size on the laser scanning device. This advantage is more significant especially in scenarios with large-sized samples.

[0058] 4. High measurement efficiency and fast speed: The swing-sweep laser scanning device proposed in the present invention has a greater advantage in information acquisition speed compared with the conventional linear scanning method. When reconstructing larger samples at a larger working distance, the information acquisition efficiency based on the swing-sweep laser scanning method is much higher than that of the translational scanning. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a structural schematic diagram of the device of the present invention;

[0060] Figure 2 is a data flow chart of the method of the present invention;

[0061] In the figure, 1 is a camera, 2 is a line laser, 3 is a high-precision rotating table, 4 is two plane mirrors, 5 is an object to be measured, 6 is a virtual image of the object to be measured formed by the mirror surface, 7 is a controller, and 8 is a computer for processing data. DETAILED DESCRIPTION

[0062] The technical solution of the present invention is further described below by means of specific embodiments in conjunction with the accompanying drawings:

[0063] Figure 1 : is a schematic diagram of the device structure of the present invention; the present invention discloses a swing-sweep laser panoramic three-dimensional scanning device based on a high-precision turntable, comprising a camera 1, a line laser 2, a rotating table 3, a controller 7, two plane mirrors 4 for multi-view imaging, and a computer 8 for processing data. The camera 1 and the line laser 2 have fixed positions and are fixed on the rotating table 3 to rotate together. A certain horizontal distance difference is required between the laser 2 and the camera 1 to generate the parallax required for three-dimensional reconstruction. The two mirrors 4 are placed behind the object to be measured and form an angle of about 120 degrees. The rotating table 3 is connected to the controller 7 via a data cable, and the computer 8 is connected to the camera and the controller at the same time, thereby realizing the synchronous control of the camera and the rotating table.

[0064] The present invention also discloses a panoramic measurement method of a swing-sweep laser panoramic three-dimensional scanning device based on a high-precision turntable. Figure 2 It is a data flow chart of the method of the present invention, comprising:

[0065] Place the calibration plate in at least 10 positions, collect the image of the calibration plate through camera 1, turn on laser 2, and shoot the laser stripe pattern projected onto the corresponding calibration plate;

[0066] Extract the coordinates of the corner points in the collected calibration image, and use Zhang's calibration method to calculate the internal and external parameters of camera 1 to obtain the plane equation of the calibration plate under different postures;

[0067] The Steger centerline extraction algorithm is used to extract the pixel coordinates on the laser stripe image. The three-dimensional coordinates of the laser line are obtained by combining the pixel coordinates of the laser line and the plane equation of the corresponding calibration plate. Finally, the three-dimensional coordinates of the laser line under N different postures are used to fit the laser plane equation.

[0068] Use camera 1 to simultaneously capture at least 5 real images of the calibration plate in different positions and the virtual image in the mirror at the initial position, and extract the 3D feature point pairs (real points, virtual points, and virtual points) of the calibration plate and the virtual image in the plane mirror 4 in the image. and virtual point Let G = {a m ,b m ,c m ,d}, we get:

[0069]

[0070] g1(G)=[1-2(a m ) 2 ]x v -2a m b m y v -2a m cm z v +2a m dx r

[0071] g2(G)=-2a m b m x v +[1-2(b m ) 2 ]y v -2b m c m z v +2b m dy r

[0072] g3(G)=-2a m c m x v -2b m c m y v +[1-2(c m ) 2 ]z v +2c m dz r

[0073] Where g1(G), g2(G), g3(G) are residual errors, N is the number of 3D point pairs; limited by the manufacturing quality of the plane mirror 4, the virtual point The accuracy cannot be guaranteed, and a systematic error is introduced into the final calibration result. A beam adjustment strategy is introduced, and the coordinates of the virtual points are optimized as variables, that is:

[0074]

[0075] Minimizing the above formula is a nonlinear minimization problem, which is solved by the Levenberg-Marquardt algorithm. By solving the above equation, the plane parameters of the plane mirror 4 can be obtained, and then the reflection matrix of the plane mirror 4 at the initial position of the camera 1 can be obtained. (in

[0076]

[0077] The camera 1 is fixed on the rotating table 3 and rotated around the rotation axis. Every time the rotating table 3 rotates around the rotation axis by a certain angle, the camera 1 collects a calibration plate image. After obtaining N calibration plate images, the external parameters of the chessboard in each image can be obtained by combining the intrinsic parameters of the camera using the PnP algorithm. Then the three-dimensional coordinates of the corner points of the chessboard in the corresponding camera coordinate system can be calculated. The relationship between the three-dimensional coordinates of a corner point of a calibration plate in the camera coordinate system of two adjacent camera positions can be expressed as:

[0078]

[0079] Among them, T is the rotation transformation matrix between the two coordinate systems, which can be expressed as:

[0080]

[0081] After obtaining N groups of rotations around the rotation axis with a rotation interval of θ c After obtaining the three-dimensional point coordinates, let G = {uvwabc}, and we can construct the equation:

[0082] G=argmin(g1(G)+g2(G)+·······+g N (G)

[0083]

[0084] Minimizing the above formula is a nonlinear minimization problem. The Levenberg-Marquardt algorithm can be used to iteratively solve the parameters of the rotation axis, and then the rotation transformation matrix of the rotation stage in the initial camera coordinate system can be calculated.

[0085] The camera coordinate system at the initial position of the rotating stage 2 is taken as the global coordinate system of the system, and the camera coordinate system at the i-th scanning position is taken as the i-th local coordinate system. The relationship between these two coordinate systems can be expressed as:

[0086]

[0087] Among them, θ i The camera is at position P i The angle of rotation from the reference position P0;

[0088] For the left plane mirror, when camera 1 is at the reference position P0, its mirror equation can be described as:

[0089]

[0090] According to the relationship between the two coordinate systems, when the camera is at position P i When , the mirror equation of the left plane mirror 4 can be described as:

[0091]

[0092] Then, in the i-th local coordinate system, the reflection matrix that transforms the point cloud coordinates of the virtual surface of the left plane mirror to the real position can be expressed as:

[0093]

[0094] in

[0095] For the right plane mirror, when camera 1 is at the reference position P0, its mirror equation can be described as:

[0096]

[0097] According to the relationship between the two coordinate systems, when the camera is at position P i When , the mirror equation of the left plane mirror 4 can be described as:

[0098]

[0099] Then, in the i-th local coordinate system, the reflection matrix that transforms the point cloud coordinates of the virtual surface of the left plane mirror to the real position can be expressed as:

[0100]

[0101] in

[0102] Turn on the rotating stage 3 and the laser 2, and use the camera 1 to sequentially collect the laser stripe patterns of the sample in the left plane mirror imaging area, the front surface area, and the right plane mirror imaging area during scanning;

[0103] Since the relative positions of the laser plane and camera 1 are fixed, when the camera position changes, the light plane equation coefficients of the laser plane do not change;

[0104] The laser stripe patterns of the left plane mirror imaging area, the front surface area and the right plane mirror imaging area of ​​the sample are converted into three-dimensional line profile information of the left plane mirror area, the front surface area and the right plane mirror area using the obtained camera 1 internal parameters and the calibrated laser plane equation;

[0105] The reflection matrix of the plane mirror 4 transforms the three-dimensional line profile information of the virtual surface of the left plane mirror area and the right plane mirror area into the real point cloud coordinates in the local camera coordinate system of the corresponding angle;

[0106] For the laser line pattern collected on the front surface of the sample, it is only necessary to transform the coordinates in the local world coordinate system into the global world system;

[0107] At this time, the three-dimensional coordinates of the sample line profile obtained are still in the local world coordinate system and can be obtained by the formula:

[0108]

[0109] The obtained three-dimensional coordinates of the virtual surface in the local coordinate system are converted into the global world coordinate system;

[0110] As the rotating stage moves, the camera collects line profile information of the object at different positions, and merges the point cloud coordinates of the camera coordinate system at the initial position obtained in different areas to obtain the panoramic three-dimensional profile information of the sample surface;

[0111] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the core technical features of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A sweeping laser panoramic 3D scanning device based on a high-precision turntable, characterized in that: The invention comprises a camera, a line laser, a rotating platform, a controller, a computer for data processing connecting the controller and the camera, and two plane mirrors for collecting multi-view images. The line laser and the camera are fixed on a rotating bracket on the rotating platform. There is a horizontal distance difference between the line laser and the camera to generate the parallax required for three-dimensional reconstruction. The two plane mirrors for collecting multi-view images are placed behind the object to be measured and form an angle. The rotating platform is connected to the controller.

2. The high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 1, characterized in that: The line laser projects a laser surface onto a plane mirror at any angle.

3. The high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 1, characterized in that: The angle between the two plane mirrors is between 90 degrees and 180 degrees.

4. The high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 3, characterized in that: The angle is 120 degrees.

5. A panoramic measurement method based on a high-precision turntable swing-sweep laser panoramic three-dimensional scanning device, characterized in that: include: Place the calibration plate in N positions, collect N pictures of the calibration plate through the camera, turn on the laser, and take N pictures of the laser stripe pattern projected onto the corresponding calibration plate; Extract the coordinates of the corner points in the collected calibration images, and use Zhang's calibration method to calculate the internal and external parameters of the camera to obtain the plane equation of the calibration plate under different postures; The Steger centerline extraction algorithm is used to extract the pixel coordinates on the laser stripe image. The three-dimensional coordinates of the laser line are obtained by combining the pixel coordinates of the laser line and the plane equation of the corresponding calibration plate. Finally, the three-dimensional coordinates of the laser line under N different postures are used to fit the laser plane equation. Fix the camera on the rotating table where the laser is located, place the calibration plate in N positions, use the camera to simultaneously shoot the calibration plate and its virtual image in the mirror, extract the three-dimensional feature point pairs of the calibration plate and the virtual calibration plate image in the mirror, and then use the Levenberg-Marquardt algorithm with bundle adjustment strategy to calculate the plane equation of the plane mirror, so as to obtain the reflection matrix of the plane mirror in the global camera coordinate system at the initial position; Turn on the rotating stage. Every time the stage rotates a certain angle around the rotation axis, the camera collects a calibration plate picture. After obtaining N calibration plate pictures, the PnP algorithm is used in combination with the camera's intrinsic parameters to obtain the extrinsic parameters of the chessboard in each picture, and the three-dimensional coordinates of the chessboard corner points in their corresponding camera coordinate system are calculated. The initial parameters of the rotating stage are calculated using the three-dimensional coordinates of the chessboard corner points at different camera positions. Finally, the rotation transformation matrix of the rotating stage in the global camera coordinate system at the initial position is obtained through optimization using the Levenberg-Marquardt algorithm. Using the plane mirror reflection matrix at the initial position and the rotation transformation matrix of the turntable, calculate the reflection matrix of the plane mirror in the local camera coordinate system when the turntable moves to different angles; Turn on the rotating stage and the laser, and use the camera to sequentially collect the laser stripe patterns of the sample in the left plane mirror imaging area, the front surface area, and the right plane mirror imaging area during scanning; The laser stripe patterns of the left plane mirror imaging area, the front surface area and the right plane mirror imaging area of ​​the sample are converted into three-dimensional line profile information of the left plane mirror area, the front surface area and the right plane mirror area using the obtained camera intrinsic parameters and the calibrated laser plane equation; The three-dimensional line profile information of the virtual surfaces of the left plane mirror area and the right plane mirror area is transformed into the real point cloud coordinates in the local camera coordinate system of the corresponding angle through the reflection matrix of the plane mirror; Use the rotation transformation matrix of the rotating stage to transform the real point cloud coordinates of the three areas in the local camera coordinate system at different angles into the point cloud coordinates in the global camera coordinate system at the initial position; The point cloud coordinates obtained in the initial camera coordinate system in different areas are merged to obtain the panoramic three-dimensional contour information of the sample surface.

6. The panoramic measurement method of the high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 5, characterized in that: The rotation transformation matrix of the turntable in the initial position global camera coordinate system is specifically: The relationship between the three-dimensional coordinates of a corner point of a calibration plate in the camera coordinate system of two adjacent cameras can be expressed as: Among them, T is the rotation transformation matrix between the two coordinate systems, expressed as: After obtaining N groups of rotations around the rotation axis with a rotation interval of θ c After obtaining the three-dimensional point coordinates, let G = {uvwabc} and construct the equation: G=argmin(g1(G)+g2(G)+······+g N (G)) Minimizing the above formula is a nonlinear minimization problem. The parameters of the rotation axis can be calculated by iteratively solving the Levenberg-Marquardt algorithm.

7. The panoramic measurement method of the high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 5, characterized in that: When the rotating stage moves to different angles, the reflection matrix of the plane mirror in the local camera coordinate system is specifically: The camera coordinate system (X0, Y0, Z0) at the initial position of the rotating stage is used as the global coordinate of the system, and the camera coordinate system (X i ,Y i ,Z i ) is taken as the i-th local coordinate system, and the relationship between the two coordinate systems can be expressed as: Among them, θ i The camera is at position P i The angle of rotation from the reference position P0, Since the relative positions of the laser plane and the camera are fixed, the light plane equation coefficients of the laser plane do not change when the camera position changes; When the camera is at the reference position P0, the mirror equation on one side is described as: According to the relationship between the two coordinate systems, when the camera is at position P i When , the mirror equation of the plane mirror can be described as: [a i b i c i d i ]=[a 0 b 0 c 0 d 0 ]T(θ i ) -1 Then, in the i-th local coordinate system, the reflection matrix that transforms the point cloud coordinates of the virtual surface of the plane mirror to the real position can be expressed as: in 8. The panoramic measurement method of the high-precision turntable-based swing-sweep laser panoramic three-dimensional scanning device according to claim 5, characterized in that: The rotation transformation matrix of the rotating stage is used to convert the real point cloud coordinates of the three areas in the local camera coordinate system at different angles into the point cloud coordinates in the global camera coordinate system at the initial position. Specifically, the camera coordinate system (X0, Y0, Z0) at the initial position of the rotating stage is used as the global coordinate of the system, and the camera coordinate system (X0, Y0, Z0) at the i-th scanning position is used as the global coordinate of the system. i ,Y i ,Z i ) is taken as the i-th local coordinate system, and the relationship between the two coordinate systems can be expressed as: Among them, θ i The camera is at position P i The angle of rotation from the reference position P0; According to the relationship between the two coordinate systems, the relationship between the local coordinates in the i-th local coordinate system and the global coordinates in the global coordinate system is: Among them, θ i The camera is located at the local coordinate position P i The angle rotated from the reference position P0 to the reference position P0.

Images