A confocal microscope camera calibration method based on high-precision motion control

Abstract

The present invention discloses a confocal microscope camera calibration method based on high-precision motion control, the main steps of which include establishing four coordinate systems, placing a checkerboard calibration plate on a high-precision platform and adjusting it to the center of the camera's field of view, and shooting and recording the image coordinates of the calibration points. Based on the imaging relationship between the optical center of the camera and the object point, the mapping relationship between the coordinate systems is determined to establish a preliminary imaging model. Subsequently, the imaging model is subjected to nonlinear correction to eliminate lens distortion, and the corrected image coordinates are converted into pixel coordinates. The nonlinear correction parameters are optimized through multi-view data, the mapping relationship between the image plane and the physical plane of the platform is further calculated, and the least squares method is used to optimize the mapping relationship to improve the calibration accuracy. Finally, the pixel coordinates in the image are converted into the physical coordinates of the platform through the optimized matrix to provide support for the autofocus process of the confocal microscope.

Description

Technical Field

[0001] The present invention relates to the technical field of microscope image processing, and in particular to a confocal microscope camera calibration method based on high-precision motion control. Background Art

[0002] Image measurement technology can be divided into two categories: two-dimensional measurement and three-dimensional measurement. Three-dimensional measurement is usually a measurement method based on stereo vision. Compared with two-dimensional measurement, it has obvious advantages in obtaining depth information. However, three-dimensional measurement requires accurate calibration of the relative positions of two or more cameras, and it is also necessary to find the correspondence between the same target point in different camera images, which involves complex stereo matching problems, thus affecting the overall measurement accuracy. In addition, there are certain difficulties in installing two cameras on the microscope platform. Since microscopes have high requirements for microscopic observation, the camera spacing needs to be kept extremely small, which puts higher requirements on the accuracy and difficulty of binocular stereo calibration. Especially when measuring tiny objects (millimeter level or even micrometer level), the use of stereo vision methods will lead to large measurement errors, and the installation of dual cameras will also bring great pressure to the spatial layout of the equipment.

[0003] For the confocal microscope platform, all its measurement parameters are located on one plane, so there is no need to use stereo vision measurement methods. If two-dimensional measurement can be combined with a high-precision motion control platform, the three-dimensional measurement problem can be converted into a two-dimensional measurement problem, thereby effectively improving the measurement accuracy. The core technical problem of two-dimensional measurement is to ensure that the optical axis of the camera maintains a perpendicular relationship with the surface of the object being measured, otherwise it will cause geometric distortion of the image and affect the accuracy of the measurement results. At present, the positioning accuracy of the motion control platform of the confocal microscope in the three directions of X, Y, and Z has reached less than 10 microns. Making full use of this precise motion control can achieve high-precision two-dimensional measurement.

[0004] like Figure 2 As shown in FIG, the confocal microscope system includes an observation camera, a measurement camera, and a laser measurement device. Figure 3 As shown in the figure, the observation camera is used to monitor the overall field of view on the platform to help users select and locate the target of interest on the platform; the measuring camera focuses on the selected target. When the measured object presents a clear image in the measuring camera, the automatic focus is completed and the focusing operation stops. According to the working distance of the measuring camera and the fixed offset relationship of the laser measuring device, the laser measuring device can be accurately controlled to focus on the measured object, thereby realizing automatic size measurement and scanning of the measured object contour.

[0005] In actual applications, when the object to be measured is at the center of the observation camera's field of view, the laser measurement device can automatically focus on the object to be measured by framing the object in the observation camera's field of view, thereby achieving contour scanning and size measurement. However, when the object to be measured is at the edge of the observation camera's field of view, even if the frame is selected, it is difficult for the laser measurement device to accurately focus on the object due to insufficient camera calibration accuracy, resulting in measurement failure. Summary of the invention

[0006] Technical purpose: In view of the insufficient calibration accuracy of existing confocal microscope cameras, the present invention discloses a confocal microscope camera calibration method based on high-precision motion control. Through high-precision camera correction technology, the calibration accuracy of the observation camera and the measurement camera is ensured, so that the camera can accurately select the target.

[0007] Technical solution: To achieve the above technical objectives, the present invention adopts the following technical solution:

[0008] A confocal microscope camera calibration method based on high-precision motion control specifically comprises the following steps:

[0009] Establish four coordinate systems, including the world coordinate system (X w , Y w , Z w ), camera coordinate system (X c , Y c , Z c ), image coordinate system (x, y) and pixel coordinate system (u, v);

[0010] Place the checkerboard calibration plate on a high-precision motion platform and ensure that the checkerboard calibration plate is located at the center of the camera's field of view and fully displayed. Take a calibration image and record the image coordinates of an object point P on the checkerboard calibration plate. c The imaging relationship with the object point P determines the mapping relationship between the coordinate systems;

[0011] Based on the image coordinates and actual coordinates of the object point P, a preliminary imaging model of the camera is established, including the intrinsic and extrinsic parameters of the camera. The intrinsic parameters include focal length, principal point coordinates and distortion coefficient, and the extrinsic parameters include rotation matrix and translation vector.

[0012] Perform nonlinear correction on the imaging model, eliminate the distortion of the camera lens by calculating the radial distortion and tangential distortion caused by the lens, and convert the image coordinates of the corrected calibration image into pixel coordinates;

[0013] Place the checkerboard calibration plate at different viewing angles, collect object point data at multiple positions, and calculate the projection difference through multi-view data to optimize the nonlinear correction parameters;

[0014] Based on the imaging model after nonlinear correction, linear correction is performed, multiple calibration points on the checkerboard calibration plate are selected, and the mapping relationship between the image plane and the physical plane of the high-precision motion platform is calculated to obtain the homography matrix H;

[0015] The homography matrix is ​​optimized using the least squares method, and the coordinate mapping of the calibration points is optimized by minimizing the back-projection error in the linear correction stage to improve the calibration accuracy.

[0016] The pixel coordinates of any object point in the image are converted into the physical coordinates of the platform through the optimized homography matrix, thereby completing the camera calibration, and the final calibration result is used for the autofocus process of the confocal microscope.

[0017] Preferably, after establishing the four coordinate systems, the camera coordinate system is normalized. The specific formula is as follows:

[0018]

[0019] Among them, x and y represent the image coordinates, X c , Y c and Z c Represents the camera coordinates.

[0020] Preferably, the preliminary imaging model of the camera is as follows:

[0021]

[0022] Among them, u and v represent pixel coordinates, x c Indicates the horizontal coordinate of the camera, x w ,y w and z w represents the world coordinates, R and T represent the orthogonal rotation transformation matrix and translation vector between the world coordinate system and the camera coordinate system, respectively, and f x =f / d x and f y =f / d y Represent the focal length components in the x and y directions, respectively, where f represents the focal length of the camera, d x and d y It represents the physical size of each pixel in the x and y axis directions, and u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

[0023] Preferably, the radial distortion and tangential distortion correction models of the camera are:

[0024]

[0025] Among them, x d and drepresents the actual coordinates of the object point P, k1, k2, k5 represent radial distortion parameters, k3, k4 represent tangential distortion parameters, and represent the tangential distortion in the x and y directions respectively.

[0026] Preferably, the specific formula for converting image coordinates to pixel coordinates is as follows:

[0027]

[0028] Among them, x d and d represents the actual coordinates of the object point P, f x =f / d x and f y =f / d y They represent the components of the focal length in the x and y directions respectively, α represents the angle between the two sides of the pixel, the angle between the actual pixel coordinate system and the image coordinate system is (π / 2-α)*180 / π, u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

[0029] Preferably, the calculation formula of the projection difference is as follows:

[0030]

[0031] Among them, proj e represents the projection difference, x i and i Represents the image coordinates of the object points under the nonlinear model at different positions of the checkerboard calibration plate, and It represents the actual image coordinates of the object point in different positions of the checkerboard calibration plate, i represents the image number of each position of the checkerboard calibration plate in the camera, and N represents the total number of positions selected by the checkerboard calibration plate.

[0032] Preferably, linear correction is performed based on the imaging model after nonlinear correction, multiple calibration points on the checkerboard calibration plate are selected, and the mapping relationship between the image plane and the physical plane of the high-precision motion platform is calculated. The specific process of obtaining the homography matrix H includes:

[0033] Select four calibration points P0, P1, P2, and P3 on the checkerboard calibration board, and record their pixel coordinates and platform physical coordinates;

[0034] The homography matrix H corresponding to the image plane coordinates and the high-precision motion platform plane coordinates is established. The relationship can be expressed as X = HX′, where X represents the image plane coordinates. X′ represents the plane coordinates of the high-precision motion platform, Homography matrix

[0035] According to the relationship between the image plane coordinates of the selected calibration points and the plane coordinates of the high-precision motion platform and the back-projection error rate calculation formula, the homography matrix H is solved by the least squares method, and the homography matrix is ​​inverted to obtain H′.

[0036] Preferably, the back-projection error rate formula is as follows:

[0037]

[0038] Among them, E reproj represents the back-projection error rate, c 00 、c 01 、c 03 、c 10 、c 11 、c 12 、c 20 、c 21 and c 22 are all elements in the homography matrix H, x j and j represents the platform physical coordinates of the calibration point, and j represents the calibration point index.

[0039] Preferably, the specific steps of converting the pixel coordinates of any object point in the image into the physical coordinates of the platform through the optimized homography matrix are as follows:

[0040] Introducing the Matrix Where H′ represents the inverse matrix of the homography matrix, u s and v s Represents the pixel coordinates of any object point;

[0041] Depend on The physical coordinates of the platform corresponding to any object point are calculated as (x s ,y s ), where x s =e0 / e2,y s =e1 / e2.

[0042] Beneficial effects: The confocal microscope camera calibration method based on high-precision motion control provided by the present invention has the following beneficial effects:

[0043] 1. The present invention ensures that the optical axis of the camera and the surface of the object to be measured are strictly perpendicular to each other by combining nonlinear correction and linear correction, eliminates radial distortion and tangential distortion of the camera lens, and eliminates possible distortion of the image at the edge, so that all pixel coordinates in the image can be accurately mapped to the physical space. Then, based on the linear mapping relationship between the image coordinates and the physical coordinates of multiple calibration points, the homography matrix is ​​calculated to achieve accurate conversion of image coordinates to physical plane coordinates. In this way, the system can control the measurement error within the micron level in a two-dimensional plane, ensure the imaging accuracy and measurement stability of the measuring camera, and greatly improve the measurement accuracy of the microscope system.

[0044] 2. The correction process of the present invention ensures the geometric consistency of the image of the observation camera in the entire field of view. Regardless of the position of the object to be measured in the field of view, accurate mapping from the image plane to the physical plane can be achieved through the homography matrix. Secondly, by utilizing the fixed relationship between the focal length of the measuring camera and the offset of the laser measuring device, the laser device can be controlled to focus on the object to be measured, and the measurement accuracy can be guaranteed even in the edge area. In this way, in the field of view of the observation camera, when the object to be measured at the edge position is framed, the system can still focus and perform contour scanning, which solves the problem of insufficient measurement accuracy of traditional confocal microscopes in edge areas and ensures the measurement reliability of the system within the entire field of view. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0046] Figure 1 It is the overall flow chart of the present invention;

[0047] Figure 2 Schematic diagram of confocal microscopy;

[0048] Figure 3 Schematic diagram of the camera image for confocal microscope observation;

[0049] Figure 4 Schematic diagram of four coordinate system models of the present invention;

[0050] Figure 5 Schematic diagram of the angle between pixel coordinates and image coordinates under the ideal model;

[0051] Figure 6 This is a schematic diagram of the angle between pixel coordinates and image coordinates in actual situations;

[0052] Figure 7 Schematic diagram of the checkerboard calibration plate in the field of view of the observation camera. DETAILED DESCRIPTION

[0053] The present invention will be described more clearly and completely below by way of a preferred embodiment in conjunction with the accompanying drawings, but the present invention is not limited to the scope of the embodiment.

[0054] like Figure 1 As shown, the present invention provides a confocal microscope camera calibration method based on high-precision motion control, which specifically includes the following steps:

[0055] S1. Establish four coordinate systems, including the world coordinate system (X w , Y w , Z w ), camera coordinate system (X c , Y c , Z c ), image coordinate system (x, y) and pixel coordinate system (u, v).

[0056] like Figure 4 As shown in the figure, it is a schematic diagram of the four coordinate system models. Specifically, the world coordinate system describes the absolute position of the calibration plate in the physical space; the camera coordinate system is based on the installation position of the camera and describes the relationship between the calibration plates; the image coordinate system uses the center of the image taken by the camera as the reference point; and the pixel coordinate system is used to describe the position of the image coordinates on the actual sensor.

[0057] In one embodiment, after establishing the four coordinate systems, the camera coordinate system is normalized. The specific formula is as follows:

[0058]

[0059] Among them, x and y represent the image coordinates, X c , Y c and Z c Represents the camera coordinates.

[0060] S2. Place the checkerboard calibration plate on a high-precision motion platform and ensure that the calibration plate is located at the center of the camera's field of view and fully displayed. Take a calibration image and record the image coordinates of an object point P on the checkerboard calibration plate. c The imaging relationship with the object point P determines the mapping relationship between the coordinate systems;

[0061] S3. Based on the image coordinates and actual coordinates of the object point P, a preliminary imaging model of the camera is established, including the intrinsic parameters and extrinsic parameters of the camera. The intrinsic parameters include focal length, principal point coordinates and distortion coefficients, and the extrinsic parameters include rotation matrix and translation vector.

[0062] In one embodiment, the preliminary imaging model of the camera is as follows:

[0063]

[0064] Among them, u and v represent pixel coordinates, x c Indicates the horizontal coordinate of the camera, x w ,y w and z w represents the world coordinates, R and T represent the orthogonal rotation transformation matrix and translation vector between the world coordinate system and the camera coordinate system, respectively, and f x =f / d x and f y =f / d y Represent the focal length components in the x and y directions, respectively, where f represents the focal length of the camera, d x and d y It represents the physical size of each pixel in the x and y axis directions, and u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

[0065] S4, performing nonlinear correction on the imaging model, eliminating the distortion of the camera lens by calculating the radial and tangential distortion caused by the lens, and converting the image coordinates of the corrected calibration image into pixel coordinates;

[0066] In one embodiment, the radial distortion and tangential distortion correction models of the camera are:

[0067]

[0068] Among them, x d and d represents the actual coordinates of the object point P, k1, k2, k5 represent radial distortion parameters, k3, k4 represent tangential distortion parameters, and represent the tangential distortion in the x and y directions respectively.

[0069] Based on the above ideal imaging model and radial distortion and tangential distortion correction models, the specific formula for converting image coordinates to pixel coordinates can be obtained as follows:

[0070]

[0071] Among them, x d and d represents the actual coordinates of the object point P, f x =f / d x and f y =f / d y They represent the components of the focal length in the x and y directions respectively, α represents the angle between the two sides of the pixel, the angle between the actual pixel coordinate system and the image coordinate system is (ππ / 2-α)*180 / π, u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

[0072] like Figure 5 and Figure 6 As shown, in the ideal model, the angle between the pixel coordinates and the image coordinates is 0, but in actual situations, α is the angle between the two sides of the pixel, which is not strictly equal to 90 degrees. The angle between the actual pixel coordinate system and the image coordinate system is (ππ / 2-α)*180 / π.

[0073] S5, placing the checkerboard calibration plate at different viewing angles, collecting object point data at multiple positions, and calculating the projection difference through the multi-view data to optimize the nonlinear correction parameters;

[0074] In one embodiment, the initial values ​​of the nonlinear distortion coefficient and α are set to 0, and the initial values ​​of the internal and external parameters of the other cameras are estimated values ​​without considering the distortion. The least squares method is used for fitting iteration to obtain the following calculation formula for the projection difference:

[0075]

[0076] Among them, proj e represents the projection difference, x i and i Represents the image coordinates of the object points under the nonlinear model at different positions of the checkerboard calibration plate, and It represents the actual image coordinates of the object point in different positions of the checkerboard calibration plate. i represents the image number of the checkerboard calibration plate in each position in the camera. N represents the total number of positions selected by the checkerboard calibration plate. The more coordinates of the checkerboard corner points, the more accurate the calibration result.

[0077] S6. Perform linear correction based on the imaging model after nonlinear correction, select multiple calibration points on the checkerboard calibration plate, calculate the mapping relationship between the image plane and the physical plane of the high-precision motion platform, and obtain the homography matrix H;

[0078] In one embodiment, linear correction is performed based on the imaging model after nonlinear correction, multiple calibration points on the checkerboard calibration plate are selected, and the mapping relationship between the image plane and the physical plane of the high-precision motion platform is calculated. The specific process of obtaining the homography matrix H includes:

[0079] Select four calibration points P0, P1, P2, and P3 on the checkerboard calibration board, and record their pixel coordinates and platform physical coordinates;

[0080] like Figure 7As shown, in one embodiment, point 0, point 10, point 77 and point 87 are selected respectively, and the pixel coordinates P0 of point 0, the pixel coordinates P1 of point 10, the pixel coordinates P2 of point 77 and the pixel coordinates P3 of point 87 are obtained through Matlab. The platform is moved so that the crosshairs in the measuring camera are aligned with the corner point centers of point 0, point 10, point 77 and point 87, and the physical coordinates PP0, PP1, PP2 and PP3 of the platform are recorded.

[0081] The homography matrix H corresponding to the image plane coordinates and the high-precision motion platform plane coordinates is established. The relationship can be expressed as X = HX′, where X represents the image plane coordinates. X′ represents the plane coordinates of the high-precision motion platform, Homography matrix

[0082] According to the relationship between the image plane coordinates of the selected calibration points and the plane coordinates of the high-precision motion platform and the back-projection error rate calculation formula, the P0-PP0, P1-PP1, P2-PP2, and P3-PP3 point pairs are brought in, and the homography matrix H is solved using the least squares method. The homography matrix is ​​then inverted to obtain H′.

[0083] The back-projection error rate formula is as follows:

[0084]

[0085] Among them, E reproj represents the back-projection error rate, c 00 、c 01 、c 03 、c 10 、c 11 、c 12 、c 20 、c 21 and c 22 are all elements in the homography matrix H, x j and j represents the platform physical coordinates of the calibration point, and j represents the calibration point index.

[0086] S7, optimizing the homography matrix using the least squares method, and optimizing the coordinate mapping of the calibration points by minimizing the back-projection error in the linear correction stage to improve the calibration accuracy;

[0087] S8. The pixel coordinates of any object point in the image are converted into the physical coordinates of the platform through the optimized homography matrix, thereby completing the camera calibration, and the final calibration result is used for the autofocus process of the confocal microscope.

[0088] In one embodiment, the specific steps of converting the pixel coordinates of any object point in the image into the physical coordinates of the platform through the optimized homography matrix are as follows:

[0089] Confirm the pixel coordinates of any object point of the object being measured and introduce the matrix Where H′ represents the inverse matrix of the homography matrix, u s and v s Represents the pixel coordinates of any object point;

[0090] Where e0 = h 00 *u s +h 01 *v s +h 02 , e1=h 10 *u s +h 11 *v s +h 12 , e2=h 20 *u s +h 21 *v s +h 22 ;

[0091] Depend on The physical coordinates of the platform corresponding to any object point are calculated as (x s ,y s ), where x s =e0 / e2,y s =e1 / e2.

[0092] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A confocal microscope camera calibration method based on high-precision motion control, characterized in that: The specific steps include: Establish four coordinate systems, including the world coordinate system (X w , Y w , Z w ), camera coordinate system (X c , Y c , Z c ), image coordinate system (x, y) and pixel coordinate system (u, v); Place the checkerboard calibration plate on a high-precision motion platform, ensure that the checkerboard calibration plate is located at the center of the camera's field of view and fully displayed, take a calibration image, record the image coordinates of an object point P on the checkerboard calibration plate, and pass through the camera optical center O c The imaging relationship with the object point P determines the mapping relationship between the coordinate systems; Based on the image coordinates and actual coordinates of the object point P, a preliminary imaging model of the camera is established, including the intrinsic and extrinsic parameters of the camera. The intrinsic parameters include focal length, principal point coordinates and distortion coefficient, and the extrinsic parameters include rotation matrix and translation vector. Perform nonlinear correction on the imaging model, eliminate the distortion of the camera lens by calculating the radial distortion and tangential distortion caused by the lens, and convert the image coordinates of the corrected calibration image into pixel coordinates; Place the checkerboard calibration plate at different viewing angles, collect object point data at multiple positions, and calculate the projection difference through multi-view data to optimize the nonlinear correction parameters; Based on the imaging model after nonlinear correction, linear correction is performed, multiple calibration points on the checkerboard calibration plate are selected, and the mapping relationship between the image plane and the physical plane of the high-precision motion platform is calculated to obtain the homography matrix H; The homography matrix is ​​optimized using the least squares method, and the coordinate mapping of the calibration points is optimized by minimizing the back-projection error in the linear correction stage to improve the calibration accuracy. The pixel coordinates of any object point in the image are converted into the physical coordinates of the platform through the optimized homography matrix, thereby completing the camera calibration, and the final calibration result is used for the autofocus process of the confocal microscope.

2. A confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: After establishing the four coordinate systems, the camera coordinate system is also normalized. The specific formula is as follows: Among them, x and y represent the image coordinates, X c , Y c and Z c Represents the camera coordinates.

3. A confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: The preliminary imaging model of the camera is as follows: Among them, u and v represent pixel coordinates, x c Indicates the horizontal coordinate of the camera, x w ,y w and z w represents the world coordinates, R and T represent the orthogonal rotation transformation matrix and translation vector between the world coordinate system and the camera coordinate system, respectively, and f x =f / d x and f y =f / d y Represent the focal length components in the x and y directions, respectively, where f represents the focal length of the camera, d x and d y It represents the physical size of each pixel in the x and y axis directions, and u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

4. A confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: The radial distortion and tangential distortion correction models of the camera are: Among them, x d and d represents the actual coordinates of the object point P, k1, k2, k5 represent radial distortion parameters, k3, k4 represent tangential distortion parameters, and represent the tangential distortion in the x and y directions respectively.

5. The confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: The specific formula for converting image coordinates to pixel coordinates is as follows: Among them, x d and d represents the actual coordinates of the object point P, f x =f / d x and f y =T / d y They represent the components of the focal length in the x and y directions respectively, α represents the angle between the two sides of the pixel, the angle between the actual pixel coordinate system and the image coordinate system is (π / 2-α)*180 / π, u0 and v0 represent the coordinates of the intersection of the camera optical axis and the image plane, that is, the principal point coordinates.

6. The confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: The projection difference is calculated as follows: Among them, proj e represents the projection difference, x i and i Represents the image coordinates of the object points under the nonlinear model at different positions of the checkerboard calibration plate, and It represents the actual image coordinates of the object point in different positions of the checkerboard calibration plate, i represents the image number of each position of the checkerboard calibration plate in the camera, and N represents the total number of positions selected by the checkerboard calibration plate.

7. The confocal microscope camera calibration method based on high-precision motion control according to claim 1, characterized in that: Based on the imaging model after nonlinear correction, linear correction is performed, multiple calibration points on the checkerboard calibration plate are selected, and the mapping relationship between the image plane and the physical plane of the high-precision motion platform is calculated. The specific process of obtaining the homography matrix H includes: Select four calibration points P0, P1, P2, and P3 on the checkerboard calibration board, and record their pixel coordinates and platform physical coordinates; The homography matrix H corresponding to the image plane coordinates and the high-precision motion platform plane coordinates is established. The relationship can be expressed as X = HX′, where X represents the image plane coordinates. X′ represents the plane coordinates of the high-precision motion platform, Homography matrix According to the relationship between the image plane coordinates of the selected calibration points and the plane coordinates of the high-precision motion platform and the back-projection error rate calculation formula, the homography matrix H is solved by the least squares method, and the homography matrix is ​​inverted to obtain H′.

8. The confocal microscope camera calibration method based on high-precision motion control according to claim 7, characterized in that: The back-projection error rate formula is as follows: Among them, E reproj represents the back-projection error rate, c 00 、c 01 、c 03 、c 10 、c 11 、c 12 、c 20 、c 21 and c 22 are all elements in the homography matrix H, x j and j represents the platform physical coordinates of the calibration point, and j represents the calibration point index.

9. A confocal microscope camera calibration method based on high-precision motion control according to claim 8, characterized in that: The specific steps to convert the pixel coordinates of any object point in the image into the physical coordinates of the platform through the optimized homography matrix are as follows: Introducing the Matrix Where H′ represents the inverse matrix of the homography matrix, u s and v s Represents the pixel coordinates of any object point; Depend on The physical coordinates of the platform corresponding to any object point are calculated as (x s ,y s ), where x s =e0 / e2,y s =e1 / e2.

Images