Binocular stereo vision calibration based on accurate ellipse detection algorithm of direct calculation and grating conversion check

Abstract

With the rapid development of science and technology, binocular vision detection technology has become one of the most important technologies because of its high precision and stability. In order to achieve high-precision three-dimensional measurement of high-speed rail train, a binocular stereo vision system is built in this paper. The visual image is gray-scale processed, and then the integral image is used to realize the edge preservation processing of local mean square deviation, and then the denoised image is enhanced. At the same time, FAST algorithm based on accelerated segmentation detection is proposed to extract feature points. In order to solve the problem of camera calibration, this paper proposes a high-precision binocular stereo vision calibration algorithm based on direct calculation and grating conversion calibration. Firstly, the reliable information points are determined from the two-dimensional image, and then the edge image is obtained by effectively detecting the ellipse contour according to the set of information points. Finally, the ellipse parameters are extracted by using the ellipse contour information, and finally the camera parameters are solved by using the ring point. Compared with the traditional calibration method, the experimental results show that the average reprojection error of the two cameras is about 0.1 pixel, and the data accuracy is high, which proves the feasibility and theoretical correctness of the experimental method.