CN-120997312-B - Calibration method and system suitable for single-camera high-precision scanning system

Abstract

The application provides a calibration method suitable for a single-camera high-precision scanning system, which comprises the following steps of S1, calibrating a linear array camera area array mode, moving a base station to image at different positions of a checkerboard by adjusting exposure time of the linear array camera, solving internal parameters and distortion parameters of the linear array camera, S2, calibrating a linear scanning mode, acquiring pixel coordinates of corner points of the checkerboard by scanning and imaging the checkerboard, separating actual rotation angles of the checkerboard from projection errors, and further pushing real rotation angles of the camera, thereby completing calibration. Aiming at the calibration requirement of the linear camera in a scanning imaging scene, the application provides a two-stage algorithm combining area array mode modeling and linear scanning imaging correction, and effectively solves the problem that the traditional calibration method is not applicable under the conditions of single-camera fixation and unidirectional movement.

Inventors

• CUI TONG
• Fang Kuai

Assignees

• 广东索鲁达科技有限公司

Dates

Publication Date

20260512

Application Date

20250928

Claims (4)

1. 1. The calibration method suitable for the single-camera high-precision scanning system is characterized by comprising the following steps of: s1, calibrating area array mode of linear array camera Moving the base station to image at different positions of the checkerboard by adjusting the exposure time of the linear array camera, and solving internal parameters and distortion parameters of the linear array camera; s2, line scanning mode calibration Obtaining pixel coordinates of corner points of the checkerboard by scanning and imaging the checkerboard lines, and separating actual rotation angles of the checkerboard from projection errors Thereby pushing the real camera rotation angle Thus, the calibration is completed; the step S1 specifically comprises the following steps S11-S14: S11, solving a joint positive matrix; S12, solving the amplification factor and the rotation angle through a positive stress matrix; S13, solving a distortion coefficient through the amplification factor; s14, carrying out iterative solution; The step S11 includes the steps of: The following linear system solutions are constructed, the rotation angle and the displacement height of the camera are unified, only different joint positive response matrixes are translated, and meanwhile, the small influence caused by the inclination angle is ignored: , Wherein the method comprises the steps of Is the coordinates of the corner points in the checkerboard coordinate system, For the pixel coordinates, To the point of The parameters are shared for the rotating part, Independently translating for each image; For the image sequence number, For the corner numbers within each image, For the height of the camera, the observation equations of a plurality of images can be spliced into a unified linear system on the right side of the matrix and the lower side of the vector, and a least square is adopted to solve a positive matrix; obtaining camera height by measuring camera base station distance, and carrying in normalized solution ; The step S12 includes: by solving out According to the unit norm constraint and the orthogonality constraint, constructing a linear system as follows: , the above is an overdetermined linear system, and the focal length magnification is linearly solved by using a least square method ; At the same time solve the rotation angle , Including the checkerboard and camera coupling rotation angle: 。
2. 2. the method of claim 1, wherein the step of determining the position of the substrate comprises, In the step S11, the image corner is subjected to anti-distortion by using the obtained internal parameters and distortion parameters, and then is solved.
3. 3. A calibration system suitable for a single camera high precision scanning system, using the method of claim 1 or 2, comprising: the linear array camera area array mode calibration module is used for moving the base station to image at different positions of the checkerboard by adjusting the exposure time of the linear array camera and solving internal parameters and distortion parameters of the linear array camera; The line scan mode calibration module is used for obtaining pixel coordinates of the corner points of the checkerboard by scanning and imaging the checkerboard lines, and separating the actual rotation angle of the checkerboard from the projection error Thereby pushing the real camera rotation angle Thus, calibration is completed.
4. 4. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor runs the computer program to implement the method of claim 1 or 2.

Description

Calibration method and system suitable for single-camera high-precision scanning system Technical Field The application relates to the technical field of camera calibration, in particular to a calibration method and a calibration system suitable for a single-camera high-precision scanning system. Background In order to achieve accurate mapping from image coordinates to spatial coordinates, any imaging-based measurement system requires geometric calibration of the camera to determine its internal, external and distortion parameters. At present, a camera calibration method is commonly used for carrying out internal reference and external reference solving in a mode of forming a regular pattern and a quantitative conversion relation, wherein a representative method is a Zhang Zhengyou method, a checkerboard or a dot array is observed through a plurality of visual angles, and internal and external parameters and distortion coefficients of a camera are solved according to homonymous points of world coordinates and image coordinates. A common feature of such methods is that multiple viewing angles are required for observation, e.g., zhang You method requires three viewing angles to resolve all internal references, while each viewing angle has an independent external reference. The calibration algorithm is widely applied to the fields of structured light measurement, stereoscopic vision, robot vision and the like, and has higher maturity. However, in high-precision scanning imaging scenarios (such as line scan imaging in wafer defect detection), its applicability is significantly inadequate, mainly in two aspects: The imaging geometry is different, namely the position and the posture of a camera in line scanning imaging are fixed, and a base station moves along an approximately horizontal plane, so that the basic assumption of the traditional method on the diversity of the external parameters cannot be satisfied, and the uncertainty of the algorithm analysis process is higher. The imaging physical process is obviously different, the line scan image is formed by pixel integration of different angles in the scanning direction, the imaging model is obviously different from a conventional area array camera, and specific effects include: 1. the imaging coordinate transformation is not rigid transformation, and the orthogonal condition of the rotation matrix array assumed by the conventional calibration algorithm is not established any more; 2. the lens distortion in the scanning direction is integrated and smoothed, while the distortion in the orthogonal direction is completely reserved, and the distortion distribution is different from the conventional model in directivity. Conventional algorithms for line scan calibration are generally only applicable to stereoscopic scan scenes with lower resolution. The single view jiao scan does not calibrate the camera internal and external parameters. Thus, such algorithms typically require additional conditions such as multiple camera line sweeps jiao degrees, three-dimensional calibration objects, planar camera assistance of known parameters, precise movement of the submount or calibration object in the vertical direction, complete forward placement of the checkerboard, etc. In the industrial high-precision scanning scene, the camera and the base station are extremely precise and fragile, and the conditions are not applicable: 1. Slight temperature and humidity changes can lead to unacceptable system errors, and touch adjustment of the optical system should be avoided as much as possible after optical correction is completed, so that scanning or vertically moving the lens by jiao degrees is not feasible. 2. The imaging resolution is usually in submicron level, so that the size of the three-dimensional calibration object needs to be in micron level to avoid defocus and punctuation errors, the process precision needs to be in submicron level, and the prior process is difficult to achieve even without cost. 3. Typically the base station cannot nor does it need to be moved in a vertical direction. The vertically movable base station cannot achieve accurate movement in submicron level. 4. In order to complete imaging in submicron level, light source and precise lens with extremely high energy level are required, and the cost is over one million. 5. The camera internal parameters can be obtained without the above-limited algorithm, such as the Donne algorithm, but the rotation jiao degrees of the checkerboard arrangement cannot be separated, namely the accurate external parameters, and the default camera movement is different from the base station, and the method is also different from the high-precision scanning scene modeling. In summary, there is no reasonable calibration algorithm for high-precision scanning systems. Therefore, the traditional area-array camera calibration method is still commonly adopted in high-resolution line scanning scenes at present. Since aberrations are