Three-dimensional tilt calibration method, device, equipment and storage medium

Abstract

The application provides a three-dimensional tilt calibration method, device, equipment and storage medium, which fuses a plane array image and a line scanning image, and constructs a space mapping relationship between multiple coordinate systems; secondly, a three-dimensional rotation matrix is introduced to explicitly represent the real attitude of a camera relative to a motion platform around an X axis, a Y axis and a Z axis; thirdly, a plane array calibration model and a line scanning calibration model are constructed in stages, while maintaining the original algorithm linear decoupling and high efficiency calculation advantages, three-dimensional tilt parameters are included in the joint optimization process, and tilt errors can be systematically corrected without additional hardware. Finally, by comprehensively integrating the first and second calibration parameters, the image re-projection error is significantly reduced, the mapping accuracy of image coordinates to space coordinates is improved, and the image reconstruction consistency and space coordinate accuracy are effectively improved in a sub-micron resolution high-precision line scanning system.

Description

Technical Field

[0001] This invention relates to the field of machine vision and image processing technology, and in particular to a method, apparatus, device and storage medium for three-dimensional tilt calibration. Background Technology

[0002] Linear scanning cameras, due to their advantages such as high resolution and high-speed imaging, are widely used in precision manufacturing scenarios such as semiconductor inspection and industrial defect identification. Unlike area scanning cameras, line scanning cameras acquire images line by line in chronological order. The construction of a complete image depends on the relative motion between the camera and the object being measured, thus requiring unique calibration methods. In existing technologies, Soluda's caliUni line scanning imaging calibration algorithm can solve for the intrinsic and extrinsic parameters of a line scanning camera at a certain resolution, making it suitable for various industrial scanning scenarios. However, this algorithm and most existing line scanning calibration methods assume that the scanning camera is strictly parallel to the motion platform, thus simplifying the extrinsic parameter modeling to a two-dimensional planar problem. It only considers the camera's rotation about the vertical axis (Z-axis) and neglects the small tilt of the camera relative to the platform about the X and Y axes. In practical systems, due to factors such as camera installation errors, lens gravitational deformation, and platform unevenness, small tilt angles about the X and Y axes are inevitably introduced. Under submicron-level high-precision scanning conditions, these micro-tilts can significantly affect the transformation between image coordinates and world coordinates, leading to image misalignment and distortion accumulation, which in turn reduces the system's reconstruction accuracy.

[0003] In summary, the problems existing in the current technology urgently need to be solved. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and storage medium for three-dimensional tilt calibration, which addresses the deficiencies in the prior art and improves the mapping accuracy from image coordinates to spatial coordinates.

[0005] This invention provides a method for three-dimensional tilt calibration, comprising: Acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan images include the coordinates of the area scan corner points, and the line scan images include the coordinates of the line scan corner points. Construct a spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship; A surface array calibration model is constructed based on the three-dimensional rotation matrix, and the first calibration parameters containing the three-dimensional rotation matrix are solved based on the coordinates of the surface array corner points. A line scan calibration model is constructed based on the first calibration parameters, and the second calibration parameters are solved based on the coordinates of the line scan corner points. The three-dimensional tilt calibration result of the imaging system is determined based on the first calibration parameter and the second calibration parameter.

[0006] According to a method for three-dimensional tilt calibration provided by the present invention, the step of constructing the spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establishing a three-dimensional rotation matrix based on the spatial mapping relationship, specifically includes: Establish the spatial mapping relationship between the calibration plate coordinate system and the motion platform coordinate system, transforming it to the camera coordinate system. A rotation parameter representing the three-dimensional attitude of the scanning camera relative to the motion platform is introduced into the spatial mapping relationship; Construct a three-dimensional rotation matrix based on the rotation parameters.

[0007] According to a method for three-dimensional tilt calibration provided by the present invention, the step of constructing a surface array calibration model based on the three-dimensional rotation matrix and solving for the first calibration parameter containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points specifically includes: Establish the projection relationship between the spatial coordinates of the calibration board and the image coordinates based on the spatial mapping relationship; The three-dimensional rotation matrix is ​​incorporated into the projection relationship to characterize the attitude change of the scanning camera relative to the motion platform; Construct a planar array calibration model based on the projection relationship; Parameter constraint relationships are constructed based on the coordinates of the area array corner points and the area array calibration model; Based on the aforementioned parameter constraints, solve for the rotation parameters and translation parameters corresponding to the three-dimensional rotation matrix; Based on the rotation parameters and the translation parameters, the magnification parameters, lens height parameters, and distortion parameters are determined to obtain the first calibration parameters.

[0008] According to a method for three-dimensional tilt calibration provided by the present invention, the step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and the translation parameter to obtain the first calibration parameter specifically includes: Based on the rotation and translation parameters, a magnification equation is constructed, and the camera magnification parameters are determined according to the magnification equation. The lens height parameter is determined based on the rotation and translation parameters. The tilt angle of the scanning camera about the X-axis and the tilt angle about the Y-axis are determined based on the rotation parameters. The normalized projection coordinates corresponding to the corner points are determined based on the magnification parameter, lens height parameter, and rotation parameter. A distortion model is constructed based on the deviation between the normalized projected coordinates and the actual image coordinates; The distortion parameters are solved based on the distortion model.

[0009] According to a method for three-dimensional tilt calibration provided by the present invention, after the step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and translation parameter to obtain the first calibration parameter, the method further includes: The distortion parameters are used to perform anti-distortion processing on the corner point coordinates of the array; The parameter constraint relationship is reconstructed based on the coordinates of the corner points of the array after anti-distortion. Re-execute the process of solving for rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters; When the parameter change or reprojection error meets the preset convergence condition, the first calibration parameter is output.

[0010] According to a three-dimensional tilt calibration method provided by the present invention, the step of constructing a line scan calibration model based on the first calibration parameters and solving for the second calibration parameters based on the coordinates of the line scan corner points specifically includes: Establish a line scan calibration model based on the first calibration parameters; Determine the corresponding normalized distortion-free coordinates based on the coordinates of the line sweep point; Construct line scan constraint equations based on the normalized distortion-free coordinates; The second calibration parameter is solved based on the line scan constraint equation.

[0011] According to a method for three-dimensional tilt calibration provided by the present invention, the step of determining the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter specifically includes: The rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters of the imaging system are determined based on the first calibration parameters. The rotation parameters are corrected according to the second calibration parameters; Based on rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters, a mapping relationship between image coordinates and target spatial coordinates is established, and the three-dimensional tilt calibration result is output.

[0012] The present invention also provides a three-dimensional tilt calibration device, comprising: The image acquisition module is used to acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan image includes the coordinates of the area scan corner points, and the line scan image includes the coordinates of the line scan corner points. The rotation matrix module is used to construct the spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and to establish a three-dimensional rotation matrix based on the spatial mapping relationship. The first calibration module is used to construct a surface array calibration model based on the three-dimensional rotation matrix, and to solve for the first calibration parameters containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points. The second calibration module is used to construct a line scan calibration model based on the first calibration parameters, and to solve for the second calibration parameters based on the coordinates of the line scan corner points. A three-dimensional calibration module is used to determine the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the three-dimensional tilt calibration method as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the three-dimensional tilt calibration method as described above.

[0015] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the three-dimensional tilt calibration method as described above.

[0016] The present invention provides a method, apparatus, device, and storage medium for three-dimensional tilt calibration. By acquiring area scan images and line scan images, and constructing spatial mapping relationships between the calibration board coordinate system, motion platform coordinate system, camera coordinate system, and image coordinate system, the present invention can fully utilize the complementary information of the two imaging modes, providing a rich data foundation for subsequent three-dimensional attitude determination. Secondly, by establishing and introducing a three-dimensional rotation matrix, the present invention overcomes the limitation of assuming strict parallelism between the camera and the platform in existing technologies, and can explicitly characterize the true attitude of the camera relative to the motion platform in three-dimensional space, including small tilts around the X and Y axes, thereby significantly improving the adaptability of the calibration model to actual installation errors and system deformation. Thirdly, the present invention constructs area scan calibration models and line scan calibration models respectively, and solves the first and second calibration parameters in stages. This maintains both a linear decoupling structure and efficient computational performance, and incorporates the three-dimensional tilt parameters into the joint optimization process, achieving systematic error correction of three-dimensional tilt without adding additional hardware. Finally, by combining the first calibration parameters and the second calibration parameters, the final tilt calibration result is determined. This invention can effectively reduce the image reprojection error caused by camera tilt, improve the mapping accuracy of image coordinates to spatial coordinates, and significantly improve the consistency of image reconstruction and spatial coordinate accuracy in high-precision line scan systems with slight tilt. It is particularly suitable for industrial linear array imaging equipment with sub-micron resolution. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the three-dimensional tilt calibration method provided by the present invention; Figure 2 This is a schematic diagram of the imaging system provided by the present invention; Figure 3 This is a schematic diagram of the structure of the three-dimensional tilt calibration device provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] To address the problems in existing technologies, this invention proposes a three-dimensional tilt calibration method to improve the mapping accuracy from image coordinates to spatial coordinates. The method is described below. Figure 1 As shown, including but not limited to the following steps: Step 110: Obtain the area scan image and the line scan image captured by the scanning camera on the calibration plate. The area scan image includes the coordinates of the area scan corner points, and the line scan image includes the coordinates of the line scan corner points.

[0021] In step 110, the calibration plate is fixed on the motion platform, while maintaining the positional relationship between the scanning camera and the motion platform. The motion platform is controlled to move, enabling the scanning camera to acquire the area scan image and line scan image corresponding to the calibration plate.

[0022] Among them, the area scan image is used to obtain the distribution information of the feature points of the calibration board in the two-dimensional image, and the line scan image is used to obtain the imaging information of the calibration board during the linear scanning process.

[0023] Furthermore, corner detection is performed on the acquired area array image to extract the area array corner coordinates corresponding to each corner point in the calibration board; corner detection is also performed on the acquired line scan image to extract the line scan corner coordinates corresponding to each corner point in the calibration board, which will be used as observation data for subsequent calibration calculations.

[0024] It should be noted that the scanning camera is a linear scan camera fixedly mounted above the motion platform. The motion platform is used to drive the calibration plate to perform displacement movement, and the calibration plate is preferably a checkerboard calibration plate. Step 120: Construct the spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship.

[0025] In step 120, the positional relationship between the calibration plate coordinate system and the motion platform coordinate system is established based on the installation position of the calibration plate on the motion platform; the positional relationship between the motion platform coordinate system and the camera coordinate system is established based on the installation relationship between the scanning camera and the motion platform; and the positional relationship between the camera coordinate system and the image coordinate system is established based on the imaging relationship of the scanning camera.

[0026] Based on this, a spatial mapping relationship is established between the calibration board coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system to describe the correspondence between the spatial position of the calibration board and the image imaging position.

[0027] Furthermore, based on the spatial attitude of the scanning camera relative to the motion platform, an attitude transformation relationship is introduced into the spatial mapping relationship to establish a three-dimensional rotation matrix for characterizing the spatial attitude of the scanning camera.

[0028] Step 130: Construct a surface array calibration model based on the three-dimensional rotation matrix, and solve for the first calibration parameters containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points.

[0029] In step 130, a surface array calibration model between the spatial coordinates of the calibration plate and the image coordinates is established based on the spatial mapping relationship and the three-dimensional rotation matrix.

[0030] The area array calibration model is used to describe the imaging process of mapping feature points of the calibration board from spatial coordinates to image coordinates, and to reflect the influence of the spatial pose of the scanning camera on the imaging results.

[0031] Subsequently, the coordinates of the corner points of the array obtained in step 110 are substituted into the array calibration model, and the first calibration parameters containing the three-dimensional rotation matrix information are obtained by solving the undetermined parameters in the model.

[0032] Step 140: Construct a line scan calibration model based on the first calibration parameters, and solve for the second calibration parameters based on the coordinates of the line scan corner points.

[0033] In step 140, a line scan calibration model is established using the first calibration parameters obtained in step 130.

[0034] The line scan calibration model is used to describe the correspondence between the spatial position of the calibration plate and the position of the line scan image during the linear scanning process of the scanning camera.

[0035] Further, the coordinates of the line scan corner points obtained in step 110 are input into the line scan calibration model, and solved in combination with the first calibration parameters to obtain the second calibration parameters used to characterize the linear scan imaging characteristics.

[0036] Step 150: Determine the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter.

[0037] In step 150, the first calibration parameters obtained in step 130 and the second calibration parameters obtained in step 140 are combined to establish the correspondence between the spatial coordinates of the imaging system and the image coordinates.

[0038] Furthermore, based on the correspondence, the spatial attitude information and imaging parameter information of the scanning camera in the motion platform are determined, thereby obtaining the three-dimensional tilt calibration result of the imaging system.

[0039] The three-dimensional tilt calibration results can be used for subsequent target measurement, size calculation, spatial positioning, and image correction applications to improve the measurement accuracy and spatial positioning accuracy of the linear array scanning imaging system.

[0040] As a further optional embodiment, the step of constructing the spatial mapping relationship between the calibration board coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establishing a three-dimensional rotation matrix based on the spatial mapping relationship, specifically includes: Establish the spatial mapping relationship between the calibration plate coordinate system and the motion platform coordinate system, transforming it to the camera coordinate system. A rotation parameter representing the three-dimensional attitude of the scanning camera relative to the motion platform is introduced into the spatial mapping relationship; Construct a three-dimensional rotation matrix based on the rotation parameters.

[0041] In this embodiment, to describe the spatial geometric relationship between the calibration plate, the motion platform, and the scanning camera, a calibration plate coordinate system, a motion platform coordinate system, a camera coordinate system, and an image coordinate system are established. The calibration plate coordinate system represents the actual spatial position of the checkerboard corner points; the motion platform coordinate system represents the position state of the calibration plate as it moves with the motion platform; the camera coordinate system represents the imaging space of the scanning camera; and the image coordinate system represents the pixel positions in the image. Through the coordinate transformation relationships between these coordinate systems, a mapping relationship between spatial points on the calibration plate and image pixels can be established, achieving a unified expression from the spatial coordinates of the calibration plate to the observation coordinates of the camera.

[0042] To characterize the actual mounting posture of the scanning camera relative to the motion platform, a three-dimensional rotation matrix is ​​introduced into the spatial mapping relationship. This three-dimensional rotation matrix consists of rotation parameters, including the tilt angle α of the scanning camera about the X-axis, the tilt angle β about the Y-axis, and the rotation angle γ about the Z-axis. The tilt angles α and β characterize the three-dimensional tilt state of the scanning camera relative to the motion platform, while the rotation angle γ characterizes the rotation state of the scanning camera about the optical axis. By introducing this three-dimensional rotation matrix, the posture deviations generated during the scanning camera mounting process can be incorporated into the calibration model, thereby improving the consistency between the calibration results and the actual imaging state.

[0043] Furthermore, to describe the spatial tilt attitude of the scanning camera relative to the motion platform, a rotation parameter characterizing the three-dimensional attitude of the scanning camera is introduced into the spatial mapping relationship. This rotation parameter characterizes the attitude deflection of the scanning camera relative to the motion platform in different directions, thereby enabling the established spatial mapping relationship to reflect the spatial geometry of the scanning camera in its actual installation state.

[0044] After obtaining the rotation parameters, a three-dimensional rotation matrix is ​​constructed based on these parameters. This three-dimensional rotation matrix characterizes the overall spatial rotation relationship of the scanning camera relative to the motion platform and serves as an attitude description parameter in the subsequent area array calibration model construction process, participating in imaging geometry modeling. By introducing this three-dimensional rotation matrix, the three-dimensional tilt attitude of the scanning camera can be incorporated into the calibration process, making the subsequently solved calibration parameters more consistent with the spatial geometric characteristics of the actual imaging system, thereby improving the accuracy of the three-dimensional tilt calibration results.

[0045] As a further optional embodiment, the step of constructing a surface array calibration model based on the three-dimensional rotation matrix and solving for the first calibration parameters containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points specifically includes: Establish the projection relationship between the spatial coordinates of the calibration board and the image coordinates based on the spatial mapping relationship; The three-dimensional rotation matrix is ​​incorporated into the projection relationship to characterize the attitude change of the scanning camera relative to the motion platform; Construct a planar array calibration model based on the projection relationship; Parameter constraint relationships are constructed based on the coordinates of the area array corner points and the area array calibration model; Based on the aforementioned parameter constraints, solve for the rotation parameters and translation parameters corresponding to the three-dimensional rotation matrix; Based on the rotation parameters and the translation parameters, the magnification parameters, lens height parameters, and distortion parameters are determined to obtain the first calibration parameters.

[0046] In this embodiment, based on the spatial mapping relationship established in step 120, a projection relationship between the spatial coordinates of the calibration board and the image coordinates is first established. The projection relationship is used to describe the imaging process of each feature point on the calibration board from its spatial position to its image position, thereby establishing the correspondence between the spatial coordinates of the calibration board and the image coordinates.

[0047] Furthermore, the three-dimensional rotation matrix established in step 120 is introduced into the projection relationship to characterize the influence of the spatial attitude change of the scanning camera relative to the motion platform on the imaging process. By introducing the three-dimensional rotation matrix as an attitude transformation parameter into the projection relationship, the established projection relationship can simultaneously reflect the geometric constraints between the spatial position of the calibration plate, the spatial attitude of the scanning camera, and the image imaging result.

[0048] Based on this, a planar array calibration model is constructed according to the projection relationship including the three-dimensional rotation matrix. The planar array calibration model is used to describe the projection relationship between the spatial coordinates of the calibration board and the coordinates of the planar array image. Based on the planar array calibration model and the coordinates of the planar array corner points, the first calibration parameters can be solved. The first calibration parameters are a set of parameters obtained in the planar array calibration stage, which include the intrinsic parameters of the scanning camera (intrinsic parameters include rotation parameters, translation parameters, magnification parameters, and lens height parameters) and distortion parameters. Among them, the rotation parameters are used to describe the spatial attitude relationship of the scanning camera relative to the motion platform; the translation parameters are used to describe the spatial positional relationship between the scanning camera and the calibration board; the magnification parameters are used to characterize the proportional relationship between the spatial size and the image size; the lens height parameters are used to characterize the distance relationship between the optical center of the scanning camera and the calibration plane; and the distortion parameters are used to characterize the radial and tangential distortion characteristics generated during lens imaging.

[0049] Subsequently, parameter constraint relationships are constructed based on the coordinates of the area array corner points obtained in step 110 and the area array calibration model. These parameter constraint relationships characterize the correspondence between the area array corner point coordinates and the parameters to be determined in the model, forming joint constraints on the model parameters by utilizing observation information from multiple area array corner points.

[0050] Furthermore, based on the parameter constraints, the rotation and translation parameters corresponding to the three-dimensional rotation matrix are solved to determine the spatial pose of the scanning camera relative to the motion platform. Then, based on the rotation and translation parameters, the magnification parameters, lens height parameters, and distortion parameters are further determined to obtain the first calibration parameters.

[0051] The first calibration parameter is used to characterize the imaging characteristics and spatial pose characteristics of the scanning camera in the area array imaging mode, and serves as the basic parameter for subsequent line scan calibration model construction and parameter solution. By using the coordinates of the area array corner points to solve the first calibration parameter, reliable initial geometric constraints can be provided for parameter calibration in the subsequent line scan mode, thereby improving the accuracy and stability of the entire three-dimensional tilt calibration process.

[0052] As a further optional embodiment, the step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and the translation parameter to obtain the first calibration parameter specifically includes: Based on the rotation and translation parameters, a magnification equation is constructed, and the camera magnification parameters are determined according to the magnification equation. The lens height parameter is determined based on the rotation and translation parameters. The tilt angle of the scanning camera about the X-axis and the tilt angle about the Y-axis are determined based on the rotation parameters. The normalized projection coordinates corresponding to the corner points are determined based on the magnification parameter, lens height parameter, and rotation parameter. A distortion model is constructed based on the deviation between the normalized projected coordinates and the actual image coordinates; The distortion parameters are solved based on the distortion model.

[0053] In this embodiment, after obtaining the rotation and translation parameters of the scanning camera relative to the motion platform, a magnification equation is first constructed based on these parameters. This magnification equation characterizes the proportional relationship between the imaging size and the actual spatial size. Solving this equation determines the magnification parameter of the scanning camera. The magnification parameter reflects the scale variation characteristics of the target object in the imaging plane, providing a scale basis for the subsequent accurate mapping between spatial coordinates and image coordinates.

[0054] Furthermore, the lens height parameter is determined based on the rotation and translation parameters. The lens height parameter characterizes the spatial relationship between the optical center of the scanning camera and the motion platform, reflecting the vertical installation state of the imaging system and thus improving the accuracy of subsequent calibration parameter calculations.

[0055] Furthermore, the tilt angles of the scanning camera around the X-axis and around the Y-axis are determined based on the rotation parameters. These tilt angles characterize the spatial tilt of the scanning camera relative to the motion platform in different directions, thereby establishing a correspondence between the actual mounting posture of the scanning camera and the imaging geometry.

[0056] After obtaining the magnification, lens height, and rotation parameters, the normalized projection coordinates corresponding to each corner point are determined based on these parameters. These normalized projection coordinates characterize the projected position of the corner point under ideal imaging conditions, thus eliminating the influence of imaging scale and spatial attitude variations on the corner point position representation.

[0057] Subsequently, a distortion model is constructed based on the deviation between the normalized projected coordinates and the actual image coordinates. This distortion model characterizes the difference between the actual imaging process and the ideal imaging process, and establishes the correspondence between image coordinate offsets and distortion parameters.

[0058] Furthermore, distortion parameters are solved based on the distortion model. These distortion parameters characterize the geometric distortion properties generated by the scanning camera lens. By solving for these distortion parameters, distortion errors in the actual imaging process can be accurately described and compensated.

[0059] Finally, the magnification parameter, lens height parameter, rotation parameter, and distortion parameter are integrated to obtain the first calibration parameter. This first calibration parameter characterizes the imaging geometry and spatial attitude characteristics of the scanning camera in area scan mode, and serves as the foundation for subsequent line scan calibration model construction and the solution of the second calibration parameter. Through this process, the corner observation information in the area scan image can be fully utilized to solve the main calibration parameters of the scanning camera, providing accurate initial geometric constraints for subsequent three-dimensional tilt calibration.

[0060] As a further optional embodiment, after the step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and translation parameter to obtain the first calibration parameter, the method further includes: The distortion parameters are used to perform anti-distortion processing on the corner point coordinates of the array; The parameter constraint relationship is reconstructed based on the coordinates of the corner points of the array after anti-distortion. Re-execute the process of solving for rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters; When the parameter change or reprojection error meets the preset convergence condition, the first calibration parameter is output.

[0061] In this embodiment, to further improve the accuracy of solving the first calibration parameter, after solving the magnification parameter, lens height parameter, and distortion parameter, the distortion parameter is used to perform anti-distortion processing on the matrix corner coordinates. Specifically, the corner positions in the matrix image are corrected according to the obtained distortion parameter to eliminate the influence of lens distortion on the corner coordinates, thereby obtaining corner coordinate data that is closer to the ideal imaging state.

[0062] After obtaining the coordinates of the corner points of the array after distortion correction, the parameter constraints are reconstructed based on the updated corner point coordinates. Since distortion correction reduces the corner point observation error, the reconstructed parameter constraints can more accurately reflect the correspondence between the model parameters and the actual observation data, thereby improving the reliability of subsequent parameter solution results.

[0063] Furthermore, the solution process for rotation, translation, magnification, lens height, and distortion parameters is re-executed using the reconstructed parameter constraints. By reusing the updated parameters in the model calculation, the previous solution results can be corrected and optimized, allowing each parameter to gradually approach its true value.

[0064] During the parameter iterative update process, the parameter results obtained from two adjacent iterations are compared and the corresponding parameter changes are calculated. At the same time, the mapping relationship between spatial points and image points is re-established based on the parameters obtained by the current solution, and the reprojection error between the predicted image coordinates and the actual image coordinates is calculated to evaluate the accuracy of the current parameter solution.

[0065] When the change in the parameter is less than a preset threshold, or the reprojection error is less than a preset error threshold, it is determined that the current parameter result has met the preset convergence condition, and the iterative calculation is stopped, and the first calibration parameter is output. If the preset convergence condition is not met, the next round of parameter solving process is performed using the updated corner coordinates and parameter constraint relationships until the convergence condition is met.

[0066] Through the above-mentioned anti-distortion correction, parameter reconstruction, and iterative optimization process, the impact of lens distortion error and parameter estimation error on the calibration results can be effectively reduced, the stability and accuracy of the first calibration parameter can be improved, and a more accurate geometric basis can be provided for the establishment of the subsequent line scan calibration model and the solution of the second calibration parameter.

[0067] As a further optional embodiment, the step of constructing a line scan calibration model based on the first calibration parameters and solving for the second calibration parameters based on the line scan corner coordinates specifically includes: Establish a line scan calibration model based on the first calibration parameters; Determine the corresponding normalized distortion-free coordinates based on the coordinates of the line sweep point; Construct line scan constraint equations based on the normalized distortion-free coordinates; The second calibration parameter is solved based on the line scan constraint equation.

[0068] In this embodiment, after obtaining the first calibration parameters, a line scan calibration model is established using the first calibration parameters. Since the first calibration parameters have already characterized the imaging geometry and spatial attitude characteristics of the scanning camera in area scan mode, the first calibration parameters can be introduced into the line scan imaging process as known conditions, thereby establishing an imaging geometry model suitable for line scan mode. The line scan calibration model is used to describe the geometric correspondence between spatial points and image points during line scan imaging. Using the coordinates of the line scan corner points in the line scan image, the second calibration parameters can be further solved. The second calibration parameters are the set of parameters obtained in the line scan calibration stage, which are used to characterize the scanning attitude characteristics during the line scan process. Preferably, the second calibration parameters include the rotation angle γ of the scanning camera around the Z-axis.

[0069] Further, the corresponding normalized distortion-free coordinates are determined based on the line scan corner point coordinates obtained in step 110. Specifically, the line scan corner point coordinates are corrected using the imaging parameters and distortion parameters in the first calibration parameters to eliminate the influence of lens distortion and imaging scale differences on the corner point position, thereby obtaining the corresponding normalized distortion-free coordinates. The normalized distortion-free coordinates can more realistically reflect the ideal projection position of the corner point during line scan imaging, providing a unified coordinate representation basis for subsequent parameter solving.

[0070] After obtaining the normalized distortion-free coordinates, line scan constraint equations are constructed based on these coordinates. These equations describe the constraint relationship between spatial geometric relationships and image observation results during line scan imaging, and establish a mathematical correlation between the parameters to be solved and the normalized distortion-free coordinates. By introducing observation information from multiple line scan corner points, joint constraints on the parameters to be solved can be formed, thereby improving the stability and reliability of the parameter solution results.

[0071] Subsequently, the second calibration parameters are solved based on the aforementioned line scan constraint equations. Specifically, based on the constraint relationships established by the normalized distortion-free coordinates, the parameters to be determined during the line scan imaging process are solved to obtain the second calibration parameters used to characterize the linear array scanning imaging characteristics. The second calibration parameters are used to supplement and improve the scanning geometric information that the first calibration parameters cannot determine, enabling the imaging model in line scan mode to fully describe the actual spatial relationship between the scanning camera and the motion platform.

[0072] Ultimately, through the synergistic effect of the first and second calibration parameters, the complete calibration of the scanning camera's three-dimensional pose and the linear array scanning imaging process is achieved, providing an accurate parameter basis for subsequent applications such as three-dimensional measurement, image correction, and spatial positioning.

[0073] As a further optional embodiment, the step of determining the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter specifically includes: The rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters of the imaging system are determined based on the first calibration parameters. The rotation parameters are corrected according to the second calibration parameters; Based on rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters, a mapping relationship between image coordinates and target spatial coordinates is established, and the three-dimensional tilt calibration result is output.

[0074] In this embodiment, after solving for the first and second calibration parameters, the rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters of the imaging system are determined based on the first calibration parameters. The rotation parameters include the tilt angle α of the scanning camera around the X-axis, the tilt angle β around the Y-axis, and the rotation angle γ around the Z-axis. The translation parameters characterize the spatial relationship between the scanning camera and the calibration plate. The magnification parameters characterize the ratio between the target spatial size and the image size. The lens height parameters characterize the distance between the optical center of the scanning camera and the calibration plane. The distortion parameters characterize the geometric distortion characteristics generated during lens imaging.

[0075] Furthermore, the second calibration parameter is a parameter obtained based on the line scan calibration model and line scan constraint equations, used to characterize the rotational attitude information of the scanning camera in the line scan imaging state. Preferably, the second calibration parameter includes the rotation angle γ of the scanning camera around the Z-axis. Since the rotation parameter in the first calibration parameter is mainly obtained based on the projection relationship established by the area array image, while the second calibration parameter is obtained by combining the actual scanning geometric constraints in the line scan image, the rotation angle γ around the Z-axis in the rotation parameter is corrected using the second calibration parameter to improve the consistency between the rotation parameter and the actual line scan imaging state.

[0076] Specifically, the tilt angle α around the X-axis and the tilt angle β around the Y-axis in the first calibration parameters are used as fixed parameters, and the rotation angle γ around the Z-axis determined by the second calibration parameters is used as the corrected rotation parameter. Together with the translation parameter, magnification parameter, lens height parameter and distortion parameter, they constitute the final calibration parameter set of the imaging system.

[0077] After obtaining the corrected rotation parameters, as well as the translation, magnification, lens height, and distortion parameters, a mapping relationship between image coordinates and target spatial coordinates is established based on these parameters. Specifically, the rotation and translation parameters establish a coordinate transformation relationship between the target spatial coordinate system and the camera coordinate system; the magnification and lens height parameters establish a projection relationship between camera coordinates and image coordinates; and the distortion parameters are used to compensate for geometric distortions during the imaging process, thereby establishing a mapping relationship between image coordinates and target spatial coordinates, and outputting the three-dimensional tilt calibration result.

[0078] In summary, a preferred embodiment of the present invention is as follows: This embodiment provides a method for three-dimensional tilt calibration, which is applied to an imaging system including a scanning camera and a motion platform.

[0079] like Figure 2As shown, firstly, the chessboard calibration board is fixed on the motion platform. The motion platform is controlled to move the calibration board, and a scanning camera is used to acquire images of the calibration board. Specifically, the scanning camera acquires both area scan images and line scan images of the calibration board. Corner points are extracted from the area scan images to obtain the coordinates of the area scan corner points; corner points are also extracted from the line scan images to obtain the coordinates of the line scan corner points.

[0080] Subsequently, a calibration plate coordinate system (X) was established. w ,Y w Z w ), motion platform coordinate system (X) b ,Y b Z b ), camera coordinate system (X) c ,Y c Z c The calibration plate coordinate system (u, v) and the image coordinate system are used. Based on the spatial relationship of the transformation from the calibration plate coordinate system to the motion platform coordinate system to the camera coordinate system, the spatial mapping relationship between the spatial points of the calibration plate and the image points is established.

[0081] The coordinates of the checkerboard corner points are transformed using two-dimensional homogeneous coordinates. Let the origin of the checkerboard coordinate system in the i-th imaging be the corner point closest to the top left corner within the field of view, and its position in the platform coordinate system be (p...). i w ,q i w Then, in the i-th image, the coordinates of the j-th corner point in the chessboard are (x...). i j ,y i j The corresponding base coordinates are ( x i Bj , y i Bj ):

[0082] Note that in actual calibration (p) i w ,q i jw It cannot be measured. It is also assumed here that the thickness of the chessboard grid is negligible and that the base is in contact with the grid without warping.

[0083] Assume that the position of the origin of the camera coordinate system in the base coordinate system during shooting is (p i b ,q i b ,Tc), where T cThis represents the camera's focusing distance. The coordinates in the camera coordinate system ( x i Cj ,y i Cj The coordinates of the chessboard square (x) can be obtained using the following formula. i j ,y i j )get:

[0084] Where the rotation matrix R A for:

[0085] Where α and β are the camera's rotation angles around the base along the X and Y axes, and γ and θ are the camera's and the checkerboard's rotation angles around the Z axis minus 90 degrees, respectively. The camera's rotation sequence is ZYX axis.

[0086] Translation vector T A for:

[0087] Since the transformation process is consistent for all i and j indices, the subscripts and superscripts are ignored in the following expressions. Projection from the camera coordinate system to the normalized plane:

[0088] The normalized plane coordinates before distortion are ( , ). Calculate the distorted coordinates (u) in the normalized plane. s ,v s ):

[0089] Where k1 and k2 are distortion parameters. Let f x ,f y c represents the focal length in the horizontal and vertical directions of the image. x ,c y The horizontal and vertical coordinates of the principal point are given. The camera image coordinates (u, v) are:

[0090] Next, we perform line scan modeling. In the line scan model, the coordinate system from the checkerboard coordinate system to the base coordinate system is the same as in the area scan model, but the coordinate system from the base coordinate system to the camera coordinate system is different. After linking, the checkerboard coordinates are (x... L j ,y L j ) to camera coordinate system coordinates ( xL Cj ,y L Cj The conversion of (superscript L indicates line scan imaging) is as follows:

[0091] Where the angle correlation matrix R L for:

[0092] Translation vector T L for:

[0093] Furthermore, in practical applications, when imaging scanning objects such as wafers, the transformation from the wafer coordinate system to the camera coordinate system is similar to that shown in this figure, taking (p i w ,q i jw The value is 0 or the preset wafer coordinate system translation parameter is based on the actual application.

[0094] Lens distortion occurs only along the X-axis of the camera and image coordinate system, and the magnification in the Y-axis can be controlled by the speed of the platform movement. (The coordinates in the camera coordinate system are...) x C ,y C Transform to the final image coordinates (u,v) (subscripts and superscripts omitted):

[0095] Next, in this stage, multiple area array images at different positions are acquired by controlling the movement of the base. The spatial-image coordinate relationship is established using the corner points of the checkerboard grid, and the camera's intrinsic and extrinsic parameters and distortion coefficients are solved.

[0096] First, we need to solve for the normalization matrix, constructing the following joint normalization matrix, where the rotation vectors are consistent across all diagrams:

[0097] Where i is the image index, j is the corner index, (x, y) are the coordinates of the corner in the checkerboard coordinate system, (u, v) are the pixel coordinates, H is the shared parameter for the rotation part, and T... i Each image is translated independently. This is because the base tilt angle is factored in. Then, to determine the liberation rate, lens height, and rotation angle, construct the following system of linear equations and solve f using least squares. x f y :

[0098] Note the assumption c x ,c y It is known that the iteration starts from 0 and is performed iteratively throughout the entire algorithm.

[0099] Get f x ,f y， c x ,c y Then, substituting into the following equation, we can solve for part of the rotation matrix:

[0100] Among them (R) 11 R 21 R 31 ) T and (R) 12 R 22 R 32 ) T Given two rotation vectors, the rotation matrix can be obtained, and the rotation angles can be solved. After simplification, the 3D rotation angles of the camera are:

[0101] Γ, the sum of the rotation angles of the base and the checkerboard, cannot be directly used for calibration. It is determined using the translation vector (T). i x ,T i y ,T i z ) T It can just eliminate the unknown quantity (p) w ,q w After simplification, we get the following concise formula:

[0102] The obtained T will then be calculated. i c Averaging to the calibrated T c As a result, the loop will converge quickly.

[0103] After solving for the magnification and rotation matrix, the deviation relationship between the normalized projection points and the real pixels is constructed, and the distortion parameters k1 and k2 are linearly estimated. Let the known normalized projection points be ( , The normalized actual image coordinates are (u / f) x ,v / f y ),but:

[0104] Finally, iterate the above steps until the result stabilizes. In solving the normality matrix, the obtained intrinsic and distortion parameters must be used to inversely distort the image corners before proceeding to achieve iterative convergence.

[0105] The line scan mode calibration is then performed. In this stage, the pixel coordinates of the corner points of the chessboard are obtained by line scan imaging of a single random track of the chessboard grid. Based on the linear scanning imaging process, the camera rotation angle γ is solved to complete the calibration.

[0106] Let the known coordinates of the checkerboard pattern in the line scan image be (x, y), and the distortion-free coordinates of the corresponding points in the image on the normalized plane be (x, y). , ); Note this ( , This is obtained by substituting image coordinates into intrinsic parameters and a distortion model. (This is similar to the area matrix calibration above, which is obtained using solved extrinsic parameters and checkerboard coordinates.) , The methods are completely different. Connect (x, y) and ( , Substituting the values ​​into the line scan model and simplifying it, we can obtain the following relationship:

[0107] Where q w γ is unknown. Let:

[0108] Substituting γ=Γ-θ, the above equation can be simplified to:

[0109] Select two points that are far apart on the line scan image (represented by subscript 12 in the following formula). We can obtain:

[0110] The three-dimensional tilt calibration device provided by the present invention will be described below, such as... Figure 3 As shown, the apparatus for three-dimensional tilt calibration described below and the method for three-dimensional tilt calibration described above can be referred to in correspondence.

[0111] A three-dimensional tilt calibration device, comprising: Image acquisition module 310 is used to acquire area scan images and line scan images captured by scanning camera on calibration plate. The area scan image includes area scan corner coordinates, and the line scan image includes line scan corner coordinates. The rotation matrix module 320 is used to construct the spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system and the image coordinate system, and to establish a three-dimensional rotation matrix based on the spatial mapping relationship. The first calibration module 330 is used to construct a surface array calibration model based on the three-dimensional rotation matrix, and to solve for the first calibration parameters containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points. The second calibration module 340 is used to construct a line scan calibration model based on the first calibration parameters and solve for the second calibration parameters based on the coordinates of the line scan corner points. The three-dimensional calibration module 350 is used to determine the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter.

[0112] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communications interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a three-dimensional tilt calibration method, which includes: Acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan images include the coordinates of the area scan corner points, and the line scan images include the coordinates of the line scan corner points. Construct a spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship; A surface array calibration model is constructed based on the three-dimensional rotation matrix, and the first calibration parameters containing the three-dimensional rotation matrix are solved based on the coordinates of the surface array corner points. A line scan calibration model is constructed based on the first calibration parameters, and the second calibration parameters are solved based on the coordinates of the line scan corner points. The three-dimensional tilt calibration result of the imaging system is determined based on the first calibration parameter and the second calibration parameter.

[0113] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program that can be stored on a non-transitory computer-readable storage medium, wherein when the computer program is executed by a processor, the computer is capable of performing the three-dimensional tilt calibration method provided by the above methods, the method comprising: Acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan images include the coordinates of the area scan corner points, and the line scan images include the coordinates of the line scan corner points. Construct a spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship; A surface array calibration model is constructed based on the three-dimensional rotation matrix, and the first calibration parameters containing the three-dimensional rotation matrix are solved based on the coordinates of the surface array corner points. A line scan calibration model is constructed based on the first calibration parameters, and the second calibration parameters are solved based on the coordinates of the line scan corner points. The three-dimensional tilt calibration result of the imaging system is determined based on the first calibration parameter and the second calibration parameter.

[0115] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for performing the three-dimensional tilt calibration provided by the methods described above, the method comprising: Acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan images include the coordinates of the area scan corner points, and the line scan images include the coordinates of the line scan corner points. Construct a spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship; A surface array calibration model is constructed based on the three-dimensional rotation matrix, and the first calibration parameters containing the three-dimensional rotation matrix are solved based on the coordinates of the surface array corner points. A line scan calibration model is constructed based on the first calibration parameters, and the second calibration parameters are solved based on the coordinates of the line scan corner points. The three-dimensional tilt calibration result of the imaging system is determined based on the first calibration parameter and the second calibration parameter.

[0116] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0117] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for three-dimensional tilt calibration, applied to an imaging system including a scanning camera and a motion platform, characterized in that, include: Acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan images include the coordinates of the area scan corner points, and the line scan images include the coordinates of the line scan corner points. Construct a spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establish a three-dimensional rotation matrix based on the spatial mapping relationship; A surface array calibration model is constructed based on the three-dimensional rotation matrix, and the first calibration parameters containing the three-dimensional rotation matrix are solved based on the coordinates of the surface array corner points. A line scan calibration model is constructed based on the first calibration parameters, and the second calibration parameters are solved based on the coordinates of the line scan corner points. The three-dimensional tilt calibration result of the imaging system is determined based on the first calibration parameter and the second calibration parameter.

2. The method for three-dimensional tilt calibration according to claim 1, characterized in that, The step of establishing the spatial mapping relationship between the calibration board coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and establishing a three-dimensional rotation matrix based on the spatial mapping relationship, specifically includes: Establish the spatial mapping relationship between the calibration plate coordinate system and the motion platform coordinate system, transforming it to the camera coordinate system. A rotation parameter representing the three-dimensional attitude of the scanning camera relative to the motion platform is introduced into the spatial mapping relationship; Construct a three-dimensional rotation matrix based on the rotation parameters.

3. The method for three-dimensional tilt calibration according to claim 1, characterized in that, The step of constructing a surface array calibration model based on the three-dimensional rotation matrix and solving for the first calibration parameter containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points specifically includes: Establish the projection relationship between the spatial coordinates of the calibration board and the image coordinates based on the spatial mapping relationship; The three-dimensional rotation matrix is ​​incorporated into the projection relationship to characterize the attitude change of the scanning camera relative to the motion platform; Construct a planar array calibration model based on the projection relationship; Parameter constraint relationships are constructed based on the coordinates of the area array corner points and the area array calibration model; Based on the aforementioned parameter constraints, solve for the rotation parameters and translation parameters corresponding to the three-dimensional rotation matrix; Based on the rotation parameters and the translation parameters, the magnification parameters, lens height parameters, and distortion parameters are determined to obtain the first calibration parameters.

4. The method for three-dimensional tilt calibration according to claim 3, characterized in that, The step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and the translation parameter to obtain the first calibration parameter specifically includes: Based on the rotation and translation parameters, a magnification equation is constructed, and the camera magnification parameters are determined according to the magnification equation. The lens height parameter is determined based on the rotation and translation parameters. The tilt angle of the scanning camera about the X-axis and the tilt angle about the Y-axis are determined based on the rotation parameters. The normalized projection coordinates corresponding to the corner points are determined based on the magnification parameter, lens height parameter, and rotation parameter. A distortion model is constructed based on the deviation between the normalized projected coordinates and the actual image coordinates; The distortion parameters are solved based on the distortion model.

5. The method for three-dimensional tilt calibration according to claim 3, characterized in that, After the step of determining the magnification parameter, lens height parameter, and distortion parameter based on the rotation parameter and translation parameter to obtain the first calibration parameter, the method further includes: The distortion parameters are used to perform anti-distortion processing on the corner point coordinates of the array; The parameter constraint relationship is reconstructed based on the coordinates of the corner points of the array after anti-distortion. Re-execute the process of solving for rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters; When the parameter change or reprojection error meets the preset convergence condition, the first calibration parameter is output.

6. The method for three-dimensional tilt calibration according to claim 1, characterized in that, The step of constructing a line scan calibration model based on the first calibration parameters and solving for the second calibration parameters based on the line scan corner coordinates specifically includes: Establish a line scan calibration model based on the first calibration parameters; Determine the corresponding normalized distortion-free coordinates based on the coordinates of the line sweep point; Construct line scan constraint equations based on the normalized distortion-free coordinates; The second calibration parameter is solved based on the line scan constraint equation.

7. The method for three-dimensional tilt calibration according to claim 1, characterized in that, The step of determining the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter specifically includes: The rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters of the imaging system are determined based on the first calibration parameters. The rotation parameters are corrected according to the second calibration parameters; Based on rotation parameters, translation parameters, magnification parameters, lens height parameters, and distortion parameters, a mapping relationship between image coordinates and target spatial coordinates is established, and the three-dimensional tilt calibration result is output.

8. A three-dimensional tilt calibration device, applied to an imaging system including a scanning camera and a motion platform, characterized in that, include: The image acquisition module is used to acquire area scan images and line scan images captured by the scanning camera on the calibration plate. The area scan image includes the coordinates of the area scan corner points, and the line scan image includes the coordinates of the line scan corner points. The rotation matrix module is used to construct the spatial mapping relationship between the calibration plate coordinate system, the motion platform coordinate system, the camera coordinate system, and the image coordinate system, and to establish a three-dimensional rotation matrix based on the spatial mapping relationship. The first calibration module is used to construct a surface array calibration model based on the three-dimensional rotation matrix, and to solve for the first calibration parameters containing the three-dimensional rotation matrix based on the coordinates of the surface array corner points. The second calibration module is used to construct a line scan calibration model based on the first calibration parameters, and to solve for the second calibration parameters based on the coordinates of the line scan corner points. A three-dimensional calibration module is used to determine the three-dimensional tilt calibration result of the imaging system based on the first calibration parameter and the second calibration parameter.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for three-dimensional tilt calibration as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for three-dimensional tilt calibration as described in any one of claims 1 to 7.

Images