A camera calibration method for large field vision metrology

Calibration is an important task in large field vision metrology. With the increase of detection range, some issues arise, such as low accuracy, excessive size of calibration target, complex operation and large computation. To address these weaknesses, a new calibration method based on two steps is presented. It is adjustment of the cameras and field calibration. The first step is to make the epipolar planes of two cameras coincident. The special spatial constraint contributes to the simplification of the calibration process. Only three independent feature points and a single image are used to complete the calibration task in second step. Therefore, the weaknesses of large size and high cost of the calibration target are solved. In addition, the settlement model abandons the traditional complex matrix computation and algorithm to simplify the calculation procedure. Finally, the result indicates that this calibration method addresses the measuring requirement of accuracy and feasibility.