A laser galvanometer scanning calibration method under a large field of view

Abstract

The application discloses a laser galvanometer scanning calibration method under a large field of view, comprising: a calibration board is designed by using a local plane mode, a plurality of calibration points and a plurality of positioning points are arranged on the calibration board, a preset number of calibration points and / or positioning points form an ideal plane, and the calibration board comprises a plurality of ideal planes; photogrammetry is performed on the calibration board, then camera calibration is performed, and world coordinates and corresponding numbers of each calibration point and positioning point on the calibration board and camera calibration parameters are obtained; laser galvanometer light plane calibration is performed according to the world coordinates and numbers of each calibration point and positioning point, the camera calibration parameters and positive and negative black and white stripe patterns projected by the laser galvanometer, and different light plane equations of the laser galvanometer under different driving voltages in a corresponding camera coordinate system are obtained. The application solves the problem that a traditional galvanometer system cannot accurately correspond galvanometer coordinates to world coordinates one by one, so that the calibration precision of the laser galvanometer system is higher, and the application scenarios are wider.

Description

Technical Field

[0001] This invention belongs to the field of laser galvanometer technology, specifically relating to a laser galvanometer scanning calibration method under a large field of view. Background Technology

[0002] Structured light 3D imaging technology, as an important branch of computer vision, has advantages such as non-contact operation and high measurement accuracy, and is widely used in industrial automation. Currently, 3D measurement systems based on Digital Light Processing (DLP) technology suffer from poor measurement accuracy, or even are unusable, in outdoor environments and large field-of-view measurement scenarios due to limitations in the brightness of the DLP light source. Compared to traditional digital light processing technology, laser galvanometers use lasers as a light source, projecting brighter stripes, exhibiting stronger resistance to ambient light interference, and are more suitable for working under large field-of-view conditions.

[0003] In laser galvanometers, taking a one-dimensional laser galvanometer system as an example, its structured light projection device consists of a laser, a micromirror, and a motor. During measurement, it boasts high scanning efficiency. The light spot emitted from the laser is projected onto the surface of the micromirror. The desired structured light pattern is scanned and projected by controlling the oscillation of the mirror and the on / off state of the laser. In a three-dimensional measurement system composed of this galvanometer module and a single camera, the spatial information of three-dimensional points is obtained through the intersection of camera rays and the light plane. Therefore, the calibration accuracy of the laser galvanometer determines the measurement accuracy of the system. During calibration, the size of the calibration plate is determined by the sensor's working field of view. In large-field-of-view scenarios, the flatness of a large-size calibration plate cannot be met, and common calibration methods cannot achieve high-precision calibration under this field of view.

[0004] One existing calibration method utilizes photosensitive sensors to calibrate a large field-of-view laser galvanometer system. Reflective points are affixed to a large field-of-view calibration plate, and the 3D coordinates of each reflective point on the calibration plate in a preset world coordinate system are determined through 3D photogrammetry. The calibration plate is then fixed on a precision guide rail. At a certain position on the guide rail, a two-dimensional laser galvanometer projects a point-like light spot. After the laser spot is projected onto the reflective target, the photosensitive sensor can quickly respond to the beam signal reflected back from the reflective target, simultaneously recording the corresponding control digital signal of the two-dimensional galvanometer scanning head. The position information of the scanned reflective target under this signal is estimated, and thus the world coordinates of the target are obtained. The digital control signals of the scanning head and the world coordinates of the reflective target are corresponding to various poses on the guide rail, and finally, the correspondence between multiple sets of two-dimensional galvanometer digital control signals and the world coordinates of spatial points is solved. Using these data as samples, the mapping relationship between the galvanometer digital control signals and the light emission vector is optimized to complete the calibration of the two-dimensional galvanometer scanning device. The problem with this method is that the precision guide rail has positioning errors, which cause errors in the world coordinates of the reflective points on the calibration plate, affecting the calibration accuracy of the galvanometer scanning device. At the same time, the two-dimensional galvanometer is a point scanning method, which has lower scanning efficiency and longer calibration data acquisition time compared with the line scanning method of the one-dimensional galvanometer, thus reducing the calibration efficiency of the galvanometer system. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides a laser galvanometer scanning calibration method with a large field of view. The technical problem to be solved by this invention is achieved through the following technical solution:

[0006] This invention provides a laser galvanometer scanning calibration method with a large field of view, comprising:

[0007] S1: A calibration plate is designed using a local plane approach. The calibration plate is provided with multiple calibration points and multiple positioning points. A preset number of calibration points and / or positioning points form an ideal plane. The calibration plate includes multiple ideal planes.

[0008] S2: Perform photogrammetry on the calibration board, then perform camera calibration to obtain the world coordinates and corresponding numbers of each calibration point and positioning point on the calibration board, as well as the camera calibration parameters;

[0009] S3: Based on the world coordinates and numbers of each calibration point and positioning point, camera calibration parameters, and the positive and negative black and white stripe patterns projected by the laser galvanometer, the optical plane of the laser galvanometer is calibrated to obtain different optical plane equations in the camera coordinate system under different driving voltages of the laser galvanometer.

[0010] In one embodiment of the present invention, the length and width of the calibration plate are respectively equal to 1 / 2 of the length and width of the maximum field of view of the laser galvanometer; the calibration points and positioning points are sub-light points, and every four calibration points or positioning points form an ideal plane, and multiple ideal planes are arranged sequentially on the calibration plate; the multiple positioning points on the calibration plate form a uniquely determined shape distribution, and under different poses, all calibration points can be determined according to the numbering of the positioning points and the calibration points and the distribution of the positioning points.

[0011] In one embodiment of the present invention, S2 includes:

[0012] S2.1: Place the compass, reference ruler, and multiple coded blocks of the photogrammetry system around the calibration plate and fix them in place;

[0013] S2.2: Take pictures of the calibration plate in blocks at different distances and in different poses, and ensure that the reference ruler and the orientation mark are within the field of view in each picture, obtain the spatial coordinates of each calibration point and positioning point on the calibration plate in the photogrammetric coordinate system, and assign different numbers to each calibration point and positioning point.

[0014] S2.3: Burn a vertical black and white stripe image and its inverse image, as well as a full white image, into the laser galvanometer module. Place the calibration plate at different working distances within the field of view. At the same working distance, change different poses to make the calibration plate image fill the entire measurement field of view and obtain the calibration plate image. The inverse image is the image of the vertical black and white stripe image after black and white are reversed. The vertical black and white stripe image and its inverse image have 256 black and white stripes and 255 stripe boundaries, respectively.

[0015] S2.4: Extract the sub-pixel coordinates of the calibration points and positioning points on each full-white image. Utilize the mapping relationship between the world coordinates and sub-pixel coordinates of the calibration points and positioning points to obtain the camera's intrinsic and extrinsic parameters, thus completing the camera calibration.

[0016] In one embodiment of the present invention, S3 includes:

[0017] S3.1: Obtain the sub-pixel coordinate values ​​of the black and white stripe boundary points of the vertical black and white stripe pattern;

[0018] S3.2: Based on the numbering information of calibration points and positioning points in the same ideal plane, camera coordinates, camera calibration results, and sub-pixel coordinates of the stripe boundary, obtain the camera coordinates of the boundary points in the ideal plane;

[0019] S3.3: Under all poses, fit the boundary light plane equation by the least squares method to the camera coordinates of the boundary points on the same boundary.

[0020] S3.4: Obtain the vector containing the laser galvanometer rotation axis using the boundary light plane equation;

[0021] S3.5: Based on the boundary light plane equations corresponding to multiple boundaries, the proportionality coefficient between the galvanometer swing angle and the voltage is obtained by fitting.

[0022] S3.6: Determine the galvanometer swing angle based on the driving voltage, determine the normal vector of the optical plane, and obtain the equation of the optical plane where any light stripe in the preset pattern is located during the laser galvanometer scanning projection process by combining the vector of the rotation axis.

[0023] In one embodiment of the present invention, S3.1 includes:

[0024] To process the vertical black and white stripe images and their inverse images collected at the same location, starting from the pixel coordinate origin of each image, different horizontal pixels in the same row of pixels are used as independent variables, and the gray values ​​corresponding to the vertical black and white stripe images and their inverse images are used as function values ​​to obtain the corresponding gray contour images.

[0025] The zero-crossing contour image is obtained by subtracting the vertical black and white stripe image and its inverse image from the two grayscale contour images. The subpixel coordinates corresponding to the zero crossings are the horizontal subpixel coordinates of the boundary points; the pixel coordinates of the boundary points in the vertical direction are the vertical pixel coordinates of the row they belong to.

[0026] In one embodiment of the present invention, S3.2 includes:

[0027] S3.21: Based on the numbering information of the calibration points and positioning points, the different ideal planes on the calibration board are distinguished. In the image data under the same pose, the three calibration points or positioning points closest to the current stripe boundary point are found using the pixel coordinates of the boundary points and calibration points / positioning points. If the three calibration points and positioning points are located on the same ideal plane, the current stripe boundary point is retained; otherwise, the current stripe boundary point is deleted.

[0028] S3.22: Based on step S3.21, determine all stripe boundary points in all images in order to obtain accurate stripe boundary points;

[0029] S3.23: Represent the spatial coordinates of each calibration point and positioning point under different poses in the camera coordinate system, solve the plane equation determined by the three nearest calibration points or positioning points of the retained boundary point in the camera coordinate system, and at the same time, obtain the camera coordinates of the current boundary point using the camera intrinsic parameters and the sub-pixel coordinates of the retained boundary points.

[0030] In one embodiment of the present invention, S3.3 includes:

[0031] Let the fitted boundary light plane equation be: Ax + By + Cz + D = 0, where C ≠ 0.

[0032] Right now: get:

[0033] ax + by + cz = 0,

[0034] Assuming there are n boundary points at different locations on the same light plane involved in the fitting, the plane to be fitted should satisfy: To minimize the partial derivatives with respect to a, b, and c, and set the derivatives to zero, solve for the parameters a, b, and c. If the calculated values ​​of a, b, and c are normal and have solutions, then it means that C is not 0, and the assumption is reasonable. Otherwise, assume that A or B is not 0, and perform the same operation to solve for a, b, and c. Finally, obtain the fitted boundary light plane equation.

[0035] In one embodiment of the present invention, S3.5 includes:

[0036] Obtain the driving voltage at a certain light stripe of the preset pattern scanned by the scanning galvanometer:

[0037]

[0038] Where, θ M U represents the maximum swing angle of the laser galvanometer. M represents the maximum driving voltage of the galvanometer, s represents the length of the preset pattern, and d represents the distance between the current light stripe and the center light stripe in the preset pattern;

[0039] The voltages U0, U1, U2, U3, ..., U at the 255 fringe boundary locations were obtained respectively. 254 ;

[0040] Based on the normal vector of the light plane of the central light stripe And the normal vectors of the other 254 boundary light planes Find the angle θ between the boundary light plane and the central light plane. i ;

[0041] The scaling factor k between the galvanometer swing angle and the voltage was obtained by fitting the voltage and included angle at the boundaries of 254 fringes.

[0042]

[0043] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0044] 1. This invention proposes a laser galvanometer scanning calibration method under a large field of view, which solves the problem that traditional galvanometer systems cannot accurately correspond the galvanometer coordinates to the world coordinates due to defects in the calibration scheme under a large field of view. This makes the laser galvanometer system more accurate in calibration and more applicable to a wider range of scenarios.

[0045] 2. Based on a one-dimensional laser galvanometer scanning system, this invention establishes a calibration model for large field-of-view scenarios. It designs a calibration plate using a local plane approach and performs calibration operations on the local plane, which greatly reduces the calibration error caused by the unevenness of the large-size calibration plate. By projecting positive and negative fringe patterns and introducing harmonic mean to extract sub-pixel coordinates of boundary points, the accuracy of obtaining high-frequency fringe boundary points is improved, thus solving the problem of low calibration accuracy of laser galvanometer systems in large field-of-view scenarios.

[0046] 3. This invention localizes the calibration plate of the large field of view, and preserves and spatially solves the boundary points of the black and white stripes projected onto the local plane in the full frame, thus overcoming the problem caused by the poor flatness of the calibration plate of the large field of view; taking the preset pattern of the laser galvanometer system as the starting point, it uses multiple light planes in the pattern for calibration, thereby improving the calibration accuracy of the galvanometer system.

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0048] Figure 1 This is a flowchart of a laser galvanometer scanning calibration method with a large field of view provided by an embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram of a calibration plate provided in an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of a single-axis galvanometer scanning calibration provided in an embodiment of the present invention;

[0051] Figure 4 This is a grayscale contour diagram provided in an embodiment of the present invention;

[0052] Figure 5 yes Figure 4 Zero-crossing contour map obtained by grayscale contour map processing;

[0053] Figure 6 This is a schematic diagram of a preset pattern scanning geometry provided in an embodiment of the present invention. Detailed Implementation

[0054] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following describes in detail a laser galvanometer scanning calibration method with a large field of view proposed according to the present invention, in conjunction with the accompanying drawings and specific embodiments.

[0055] The foregoing and other technical contents, features, and effects of the present invention will be clearly presented in the following detailed description of specific embodiments in conjunction with the accompanying drawings. Through the description of the specific embodiments, a more in-depth and concrete understanding can be gained of the technical means and effects adopted by the present invention to achieve its intended purpose. However, the accompanying drawings are for reference and illustration only and are not intended to limit the technical solutions of the present invention.

[0056] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations are intended to cover non-exclusive inclusion, such that an article or apparatus comprising a list of elements includes not only those elements but also other elements not expressly listed. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or apparatus that includes said element.

[0057] Please see Figure 1 , Figure 1 This is a flowchart of a laser galvanometer scanning calibration method under a large field of view, provided by an embodiment of the present invention. The laser galvanometer scanning calibration method includes the following steps:

[0058] S1: A calibration plate is designed using a local plane approach. The calibration plate is provided with multiple calibration points and multiple positioning points. A preset number of calibration points and / or positioning points form an ideal plane. The calibration plate includes multiple ideal planes.

[0059] In this embodiment, the area of ​​the calibration plate is equal to 1 / 4 of the maximum measurement field of view of the laser galvanometer; that is, the length and width of the calibration plate are both equal to 1 / 2 of the length and width of the maximum field of view of the laser galvanometer. The calibration plate has multiple calibration points and multiple positioning points. Please refer to [link to relevant documentation]. Figure 2 , Figure 2 This is a schematic diagram of a calibration plate provided in an embodiment of the present invention. The calibration points and positioning points are both sub-light points, and their size is determined by the maximum field of view. Every four points (calibration points or positioning points) form a small plane, i.e., an ideal plane, and multiple small ideal planes are arranged sequentially on the calibration plate.

[0060] In actual measurement, the overall flatness of the calibration plate cannot be guaranteed, but it is assumed that the small plane formed by the above four points is an ideal plane. Positioning points are used to determine the location of calibration points during the calibration process. Multiple positioning points form a uniquely defined shape distribution on the calibration plate, with dimensions larger than the calibration points. Under different poses, based on the numbering of the positioning and calibration points, the distribution of the positioning points can determine all calibration points. In other words, positioning points are markers that distinguish ordinary calibration points, enabling image acquisition from different poses, and the coordinate systems of each calibration point and positioning point can correspond one-to-one. Specifically, the four calibration points and / or positioning points are located at the four corners of a small square. The entire calibration plate is composed of multiple small squares, each forming a small ideal plane.

[0061] S2: Perform photogrammetry on the calibration board, then complete the camera calibration, and obtain the world coordinates and corresponding numbers of each calibration point and point location on the calibration board, as well as the camera calibration parameters.

[0062] Step S2 in this embodiment specifically includes:

[0063] S2.1: Place and fix the compass rose and reference ruler of the photogrammetric system around the calibration plate. The compass rose is the photogrammetric coordinate system through which the world coordinates of each calibration point and positioning point on the calibration plate are established.

[0064] Please see Figure 3 , Figure 3 This is a schematic diagram of a single-axis galvanometer scanning calibration provided in an embodiment of the present invention. The photogrammetry system used in this embodiment is the V-stars system (Industrial Digital Close-Range Photogrammetric Coordinate Measuring System). Specifically, the calibration plate is stably placed on the floor of the experimental environment, and the characteristic components of the V-stars system, such as the compass rose, reference ruler, and coding block, are placed around the calibration plate and fixed with adhesive. The reference ruler is made of carbon fiber, which has minimal thermal expansion and contraction due to temperature changes; its length can be considered constant during measurement in different environments. It is used to constrain dimensions in photogrammetry (because the field of view varies at different shooting distances, the reference ruler can serve as a reference for scene dimensions).

[0065] S2.2: Take pictures of the calibration plate in blocks at different distances and in different poses, and ensure that the reference ruler and the orientation mark are within the field of view in each picture, obtain the spatial coordinates of each calibration point and positioning point on the calibration plate in the photogrammetric coordinate system, and assign different numbers to each calibration point and positioning point.

[0066] Due to the large size of the calibration board, this embodiment involves taking pictures of the calibration board in sections. The multiple sets of pictures are then imported into the V-stars system software. By using the multi-view geometric relationship between the images, the spatial coordinates of each calibration point and positioning point on the calibration board in the standard coordinate system, i.e., the world coordinates, are solved. At the same time, each calibration point and positioning point is assigned a number on the calibration board.

[0067] Specifically, photogrammetric images of the calibration board are captured at different distances from the camera. At the same distance, by adjusting the camera's shooting area and pose, the compass rose and reference ruler are kept within the field of view for each shot, ultimately acquiring all calibration points and positioning points on the calibration board at that distance. It should be noted that the V-stars system has a built-in compass rose (a photogrammetric coordinate system), which is fixed next to the calibration board and serves as the world coordinate system for the calibration and positioning points.

[0068] S2.3: A vertical black-and-white stripe pattern and its inverse, as well as a completely white image, are burned into the laser galvanometer module. Within the field of view, the calibration board is placed at different working distances. At the same working distance, different poses are changed to allow the calibration board image to fill the entire measurement field of view, and the image is acquired. In this embodiment, the vertical black-and-white stripe pattern, its inverse, and the completely white image are all preset patterns.

[0069] Specifically, a high-frequency vertical black-and-white stripe pattern and its inverse (i.e., the original black-and-white stripe pattern reversed) along with a completely white image are burned into the laser galvanometer module. Within the field of view, the calibration board is placed at different working distances. At the same working distance, different poses and positions are varied to ensure the calibration board image fills the entire measurement field of view. To prevent ambient light interference, a filter corresponding to the laser wavelength is added to the camera lens. Simultaneously, the laser brightness and camera exposure time are adjusted to achieve the optimal image quality, completing one pose acquisition. Finally, images of the calibration board at different working distances and poses are obtained. The working distance is the distance from the object (calibration board) to the laser galvanometer system.

[0070] In this embodiment, the high-frequency vertical black and white stripe pattern includes 256 black and white stripes, with a total of 255 boundaries. Generally, the more boundaries there are, the more data can be used for fitting and calculation, resulting in better optimization results.

[0071] S2.4: Extract the subpixel coordinates of the calibration points and localization points contained in each full white image, and use the mapping relationship between the world coordinates and subpixel coordinates of the calibration points and localization points to obtain the camera's intrinsic and extrinsic parameters, thus completing the camera calibration.

[0072] In this step, the sub-pixel coordinates of the calibration points and positioning points contained in each full-white image are extracted. Based on Zhang's calibration method, camera calibration is completed, and the calibration result satisfies the following equation:

[0073]

[0074] Where ρ is the scaling factor, represents the component of each calibration point or positioning point on the calibration board in the Z direction of the camera coordinate system, u and v represent the sub-pixel coordinates of a calibration point or positioning point, and X... W Y W Z W Let f, dx, dy, u0, and v0 represent the world coordinates of a calibration point or positioning point, respectively. f, dx, dy, u0, and v0 are the camera intrinsic parameters to be determined. f represents the camera focal length, dx and dy are the dimensions of a pixel in the u and v directions of pixel coordinates, respectively, u0 and v0 represent the pixel coordinates of the center of the camera target surface, and R and T are the rotation and translation matrices to be determined in the camera extrinsic parameter matrix.

[0075] S3: Based on the world coordinates of each calibration point and positioning point, the camera calibration parameters, and the positive and negative black and white stripe patterns projected by the laser galvanometer, the optical plane of the laser galvanometer is calibrated to obtain the different optical plane equations in the camera coordinate system under different driving voltages of the laser galvanometer.

[0076] In this embodiment, step S3 includes:

[0077] S3.1: Obtain the sub-pixel coordinates of the black and white stripe boundary points of the vertical black and white stripe pattern.

[0078] Specifically, the vertical black-and-white stripe image and its inverse image, acquired from the same location, are processed by starting from the pixel coordinate origin of each image. Within the same row of pixels, different horizontal pixels are used as independent variables, and the corresponding grayscale values ​​of the vertical black-and-white stripe image and its inverse image are used as function values ​​to create corresponding grayscale contour images, such as... Figure 4 As shown, it should be noted that the origin of pixel coordinates is generally the center of the pixel at the top left corner of the image.

[0079] Subsequently, the difference between the two grayscale contour images—the vertical black-and-white stripe image and its inverse image—is calculated (e.g., blue minus red in the grayscale contour image) to obtain a zero-crossing contour image, as shown below. Figure 5 As shown, the subpixel coordinates corresponding to the zero-crossing points are the horizontal subpixel coordinates of the boundary points. Here, the boundary points refer to the black and white stripe boundaries that are represented as points on the image pixels, and their pixel coordinates need to be solved.

[0080] Figure 5 The red dashed line in the zero-crossing contour map represents the 0-value line. In the map, a and b are the adjacent integer pixel values ​​in the horizontal direction of the boundary point, respectively. After subtracting the front and back images (vertical black and white stripe image and its reverse image), if the grayscale values ​​corresponding to pixels a and b are k1 and k2 respectively, the sub-pixel coordinates of the boundary point in the horizontal direction can be calculated using the harmonic mean:

[0081]

[0082] The pixel coordinates in the vertical direction are the vertical pixel coordinates of the row. Similarly, the sub-pixel coordinates of all boundary points in the stripe pattern can be calculated.

[0083] S3.2: Based on the numbering information of calibration points and positioning points within the same small plane, camera coordinates, camera calibration results, and sub-pixel coordinates of the stripe boundary, obtain the camera coordinates of the boundary points within the small plane.

[0084] Specifically, step S3.2 of this embodiment includes:

[0085] S3.21: Based on the numbering information of calibration points, positioning points and boundary points, different small planes (i.e. ideal planes) on the calibration plate can be distinguished. In the image data of the same pose, the three calibration points or positioning points closest to the stripe boundary point are found by using the pixel coordinates of the boundary point, calibration point and positioning point. If these three points are exactly on the same small plane, the stripe boundary point information is retained; otherwise, the stripe boundary point is deleted, thereby obtaining the accurate stripe boundary point.

[0086] S3.22: Based on step S3.21, all boundary points in all images are judged to obtain accurate boundary points.

[0087] S3.23: After the camera calibration is completed, the world coordinates of each calibration point and positioning point under different poses are expressed in the camera coordinate system. The plane equations determined by the three nearest calibration points or positioning points of the retained boundary point in the camera coordinate system are solved. At the same time, the camera coordinates of the boundary point are obtained by using the camera intrinsic parameters and the sub-pixel coordinates of the retained boundary point.

[0088] S3.3: Under all poses, project the same fringe boundary onto the camera coordinates of the fringe boundary points on the calibration plate, and fit the boundary light plane equation using the least squares method.

[0089] Specifically, let the fitted boundary light plane equation (i.e., the light plane equation at the light stripe boundary) be: Ax + By + Cz + D = 0, where C ≠ 0, that is:

[0090] We get ax + by + cz = 0.

[0091] Assuming there are n spatial points (the collective term for boundary points at different locations on the same light plane) involved in the fitting, the plane to be fitted should satisfy: Minimum.

[0092] Take the partial derivatives with respect to a, b, and c, and set the derivatives to zero, as follows:

[0093]

[0094] Then:

[0095]

[0096] Express the above equation in matrix form and solve for parameters a, b, and c:

[0097]

[0098] If the calculated values ​​of a, b, and c are normal and have solutions, it means that C is not 0, and the assumption is reasonable. Otherwise, follow the same method, assume that A or B is not 0, perform the corresponding operations to solve for a, b, and c, and finally obtain the fitted boundary light plane equation.

[0099] S3.4: Use the boundary light plane equation to obtain the vector of the rotation axis of the laser galvanometer.

[0100] All light planes emitted from the laser galvanometer pass through the galvanometer's rotation axis. The equation of the light plane can be determined by the normal vector of the light plane and the vector containing the rotation axis. Let the direction vector of the galvanometer's rotation axis be (lx, ly, lz), and let the galvanometer's rotation axis pass through the point (x0, y0, z0). Through the light plane fitting step in S3.3, 255 boundary light plane equations are obtained. The normal vector of each light fringe boundary surface is perpendicular to the rotation axis direction vector, and the points on the galvanometer's rotation axis lie on each light fringe boundary surface. Using the least squares method (similar to the approach of fitting the light plane), the vector containing the laser galvanometer's rotation axis is calculated. To facilitate the subsequent calculation of the optical plane normal vector during the laser galvanometer's oscillation process, its direction is defined as from top to bottom along the axis of rotation.

[0101] S3.5: Based on the boundary light plane equations corresponding to multiple boundaries, the proportionality coefficient between the galvanometer swing angle and the voltage is obtained by fitting.

[0102] Specifically, the swing angle of the laser galvanometer is controlled by voltage, and its swing angle changes linearly with the voltage, such as: θ = kU, where k is a proportionality coefficient, θ represents the swing angle of the laser galvanometer, and U represents the voltage. In actual use, the voltage is ideally given according to the swing angle in the preset pattern (the preset pattern here is the front and back images and the all-white image of the high-frequency longitudinal black and white stripes mentioned above). Due to errors in the motion characteristics, the actual swing angle has a slight uncertainty in the proportionality coefficient k. Based on the boundary light plane equations corresponding to multiple boundaries obtained, this invention fits the proportionality coefficient k between the laser swing angle and the voltage. The specific steps are as follows:

[0103] Voltage in ±U M Drive the galvanometer to rotate ±θ within the range M The entire image is scanned using an angle. Assuming the length of the preset pattern is s, and the distance from a certain light bar in the preset pattern to the center light bar is d, please refer to [link / reference]. Figure 6 , Figure 6 This is a schematic diagram of a preset pattern scanning geometry provided in an embodiment of the present invention (this schematic diagram is a top view of galvanometer scanning). Figure 6 For the ideal galvanometer scan image of the preset pattern, point O represents the position of the galvanometer axis; line AB represents the length of the preset pattern, with a value of s; line OC represents the distance from the center of the light stripe (i.e., the light stripe corresponding to the center of the preset pattern) to the galvanometer scanning axis, with a value of D; point F represents the position of a certain light stripe, at which the galvanometer swing angle is α; line CF represents the distance of the light stripe from the center light stripe, with a value of d.

[0104] have to:

[0105] The magnitude of the driving voltage for the laser galvanometer to scan the light stripe is:

[0106]

[0107] Among them, U M θ represents the maximum driving voltage of the galvanometer. M This indicates the maximum swing angle of the laser galvanometer, which is an acute angle; the scanning light plane position is negative to the left of the center position and positive to the right.

[0108] That is, during the calibration process, the voltages U0, U1, U2, U3...U at 255 boundary locations. 254 To achieve what is desired.

[0109] Given that the normal vector of the light plane of the central light stripe is If the driving voltage is U0 = 0, then the galvanometer swing angle θ0 = 0, based on the normal vectors of the other 254 boundary light planes. The angle between a given light plane and its initial position can be determined:

[0110]

[0111] Furthermore, the sign of the swing angle is determined based on the position of the light plane.

[0112] θ i =kU i When fitting k among 254 angles, the following should be satisfied:

[0113]

[0114] Take the derivative with respect to the parameter k and set it to 0:

[0115]

[0116] In the above formula, k represents the fitted result:

[0117]

[0118] S3.6: Based on the vector of the rotation axis of the laser galvanometer and the normal vector of the initial position light plane, obtain the equation of the light plane where any light stripe of the preset pattern is located during the laser galvanometer scanning and projection process.

[0119] As can be seen from the above steps, in the preset pattern, if the position of any light stripe in the preset pattern is known, the driving voltage U can be calculated, and then the swing angle θ = kU of the galvanometer at that light stripe can be calculated.

[0120] The vector containing the mirror rotation axis is: The normal vector of the initial position light plane is: The normal vector of the light plane after swinging by an angle θ is:

[0121]

[0122] The initial position is the position when the galvanometer swing angle is zero degrees.

[0123] The normal vector and the vector containing the rotation axis can determine the light plane. Therefore, during the scanning and projection process of the galvanometer, the equation of the light plane containing any light stripe in the preset pattern can be expressed in the camera coordinate system, and the light plane calibration of the laser galvanometer system is completed.

[0124] This invention proposes a laser galvanometer scanning calibration method under a large field of view, which solves the problem that traditional galvanometer systems cannot accurately correspond the galvanometer coordinates to the world coordinates due to defects in the calibration scheme under large field of view. This results in higher calibration accuracy of the laser galvanometer system and a wider range of applicable scenarios.

[0125] This invention, based on a one-dimensional laser galvanometer scanning system, establishes a calibration model for large field-of-view scenarios. It utilizes a local plane approach to design the calibration plate and performs calibration operations on this local plane, significantly reducing calibration errors introduced by the unevenness of the large-size calibration plate. By projecting positive and negative fringe patterns and introducing harmonic mean to extract sub-pixel coordinates of boundary points, the accuracy of acquiring high-frequency fringe boundary points is improved, solving the problem of low calibration accuracy of laser galvanometer systems in large field-of-view scenarios. This invention localizes the calibration plate for large field-of-view scenarios, preserving and spatially solving the black and white fringe boundary points projected onto the local plane from the full-frame image, overcoming the problem of poor flatness of the calibration plate in large field-of-view scenarios. Using the preset pattern of the laser galvanometer system as a starting point, it achieves optical plane calibration at any position within the pattern, improving the efficiency and accuracy of subsequent measurements.

[0126] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A laser galvanometer scanning calibration method with a large field of view, characterized in that, include: S1: A calibration plate is designed using a local planar approach. The calibration plate has multiple calibration points and multiple positioning points. A preset number of calibration points and / or positioning points form an ideal plane. The calibration plate includes multiple ideal planes. The length and width of the calibration plate are respectively equal to half the length and width of the maximum field of view of the laser galvanometer. The calibration points and positioning points are sub-light points. Every four calibration points and / or positioning points form an ideal plane. Multiple ideal planes are arranged sequentially on the calibration plate. The multiple positioning points on the calibration plate form a uniquely determined shape distribution. Under different poses, all calibration points can be determined based on the numbering of the positioning points and the calibration points, as well as the distribution of the positioning points. S2: Perform photogrammetry on the calibration board, then perform camera calibration to obtain the world coordinates and corresponding numbers of each calibration point and positioning point on the calibration board, as well as the camera calibration parameters; S3: Based on the world coordinates of each calibration point and positioning point, camera calibration parameters, and the positive and negative black and white stripe patterns projected by the laser galvanometer, the optical plane of the laser galvanometer is calibrated to obtain different optical plane equations in the camera coordinate system under different driving voltages of the laser galvanometer.

2. The laser galvanometer scanning calibration method under a large field of view according to claim 1, characterized in that, S2 includes: S2.1: Place the compass, reference ruler, and multiple coded blocks of the photogrammetry system around the calibration plate and fix them in place; S2.2: Take pictures of the calibration plate in blocks at different distances and in different poses, and ensure that the reference ruler and the orientation mark are within the field of view in each picture, so as to obtain the spatial coordinates of each calibration point and each positioning point on the calibration plate in the photogrammetric coordinate system, and assign different numbers to each calibration point and positioning point. S2.3: Burn a vertical black and white stripe image and its inverse image, as well as a full white image, into the laser galvanometer module. Place the calibration plate at different working distances within the field of view. At the same working distance, change different poses to make the calibration plate image fill the entire measurement field of view and obtain the calibration plate image. The inverse image is the image of the vertical black and white stripe image after black and white are reversed. The vertical black and white stripe image and its inverse image have 256 black and white stripes and 255 stripe boundaries, respectively. S2.4: Extract the subpixel coordinates of the calibration points and localization points on each full-white image. Utilize the mapping relationship between the world coordinates and subpixel coordinates of the calibration points and localization points to obtain the camera's intrinsic and extrinsic parameters, thus completing the camera calibration.

3. The laser galvanometer scanning calibration method under a large field of view according to claim 2, characterized in that, S3 includes: S3.1: Obtain the sub-pixel coordinate values ​​of the black and white stripe boundary points of the vertical black and white stripe pattern; S3.2: Based on the numbering information of calibration points and positioning points in the same ideal plane, camera coordinates, camera calibration results, and sub-pixel coordinates of the stripe boundary, obtain the camera coordinates of the boundary points in the ideal plane; S3.3: Under all poses, fit the boundary light plane equation by the least squares method to the camera coordinates of the boundary points on the same boundary. S3.4: Obtain the vector containing the rotation axis of the laser galvanometer using the boundary light plane equation; S3.5: Based on the boundary light plane equations corresponding to multiple boundaries, the proportionality coefficient between the galvanometer swing angle and the voltage is obtained by fitting. S3.6: Determine the galvanometer swing angle based on the driving voltage, determine the normal vector of the optical plane, and obtain the equation of the optical plane where any light stripe of the preset pattern is located during the laser galvanometer scanning projection process by combining the vector of the rotation axis.

4. The laser galvanometer scanning calibration method under a large field of view according to claim 3, characterized in that, S3.1 includes: To process the vertical black and white stripe images and their inverse images collected at the same location, starting from the pixel coordinate origin of each image, different horizontal pixels in the same row of pixels are used as independent variables, and the gray values ​​corresponding to the vertical black and white stripe images and their inverse images are used as function values ​​to obtain the corresponding gray contour images. The zero-crossing contour image is obtained by subtracting the vertical black and white stripe image and its inverse image from the two grayscale contour images. The subpixel coordinates corresponding to the zero crossings are the horizontal subpixel coordinates of the boundary points; the pixel coordinates of the boundary points in the vertical direction are the vertical pixel coordinates of the row they belong to.

5. The laser galvanometer scanning calibration method with a large field of view according to claim 3, characterized in that, S3.2 includes: S3.21: Based on the numbering information of the calibration points and positioning points, the different ideal planes on the calibration board are distinguished. In the image data under the same pose, the three calibration points or positioning points closest to the current stripe boundary point are found using the pixel coordinates of the boundary points and calibration points / positioning points. If the three calibration points or positioning points are located on the same ideal plane, the current stripe boundary point is retained; otherwise, the current stripe boundary point is deleted. S3.22: Based on step S3.21, determine all stripe boundary points in all images to obtain accurate stripe boundary points; S3.23: Represent the spatial coordinates of each calibration point and positioning point under different poses in the camera coordinate system, solve the plane equation determined by the three nearest calibration points or positioning points of the retained boundary point in the camera coordinate system, and at the same time, obtain the camera coordinates of the current boundary point using the camera intrinsic parameters and the sub-pixel coordinates of the retained boundary points.

6. The laser galvanometer scanning calibration method under a large field of view according to claim 3, characterized in that, S3.3 includes: Let the fitted boundary light plane equation be: Let among them C ≠0; Right now: ,make , , ,get: ， Assume there are a total of boundary points at different locations on the same light plane involved in the fitting. n The plane to be fitted should satisfy: Minimum, for a , b , c Find the partial derivatives separately, and set the derivatives to zero to solve for the parameters. a , b , c If the result is a , b , c If the value is normal and has a solution, it means that C is not 0, and the assumption is reasonable. Otherwise, assume that A or B is not 0, perform the same operation to solve for a, b, c, and finally obtain the fitted boundary light plane equation.

7. The laser galvanometer scanning calibration method with a large field of view according to any one of claims 3 to 6, characterized in that, S3.5 includes: Obtain the driving voltage at a certain light stripe of the preset pattern scanned by the laser galvanometer: ， in, This indicates the maximum swing angle of the laser galvanometer. represents the maximum driving voltage of the galvanometer, s represents the length of the preset pattern, and d represents the distance between the current light stripe and the center light stripe in the preset pattern; The voltage at each of the 255 fringe boundary locations was obtained. ; Based on the normal vector of the light plane of the central light stripe And the normal vectors of the other 254 boundary light planes Find the angle between the boundary light plane and the central light plane. ; The scaling factor k between the galvanometer swing angle and the voltage was obtained by fitting the voltage and included angle at the boundaries of 254 stripes. 。

Images