Target control point position calibration based on coordinate measuring machine and global bundling optimization algorithm
Min Zhao,Xing Ma,Gang Yi Hu
DOI: https://doi.org/10.1063/5.0219276
2024-10-01
Abstract:The target-based coordinate measurement system is primarily composed of a target, a camera, and a computer, relying on image processing to achieve coordinate measurement. It finds extensive applications in spatial positioning and precision measurement, such as industrial manufacturing, robot navigation, and shipbuilding. The accuracy of the target-based coordinate measurement system is highly dependent on the calibration precision of control point positions. A new method based on the calibration of control point positions using a traditional coordinate measuring machine is proposed in this paper. This method does not require the individual calculation of spatial positions for each control point; instead, it utilizes a global bundle algorithm to obtain the positions of all the control points within the field of view, thereby simplifying the calibration process. The simulation results indicate that the key factor influencing the accuracy of visual calibration is the motion range of the measured target within the field of view, and it is positively correlated with calibration accuracy. For a large-sized target, a method of stitching control points within a near-sighted distance is proposed, which achieves higher accuracy compared to full-field calibration without requiring all the control points to be present within the field of view. Experiments demonstrate that using the proposed calibration method, the standard deviations of measurement accuracy of the target-based coordinate measurement system in the x, y, and z directions are 0.01, 0.01, and 0.08 mm, respectively.