Research On Calibration Technology Applied In Stereo Robot Vision

With the development of computer science, people want the computer to be able to process information as fast as humans, so the emerging research topic of robot stereo vision was born. The main task of the 3D reconstruction is how to quickly restore the two-dimensional image information into three-dimensional space. It is one of the research directions of the robot stereo vision and the camera calibration technology is an indispensable step in this research direction. Camera calibration technology is one of most important technologies in robot stereo vision technology. The stability of the camera calibration algorithm and the accuracy of the calibration results have a direct impact on the accuracy of the robot stereo vision system. In this paper, camera calibration technology applied in the robot vision has been studied. The research focus of this paper is extraction of image features and acquisition of calibration parameters in the calibration process. The main research contents and innovative research work in this paper are as follows:1) From the beginning of data collection and literature review, this paper makes a research on the camera imaging model and the projection transformation in the process of camera imaging. In order to improve the peak signal to noise ratio of the image, the image collected by the camera is used to balance the image and remove the noise. The processed image has an important influence on the accuracy of the camera calibration.2) A corner detection algorithm based on the checker is proposed in this paper. The algorithm can shield the fillet effect caused by the Gaussian convolution in the checkerboard corner detection, even without considering the shape of the target area when the template moves. The experimental results show that this algorithm has good detection effect on the corner of the board.3) Double vanishing point calibration method based on the checker is proposed in this paper. The calibration of camera focal length is carried out by two vanishing points of two parallel lines, which are orthogonal to each other. Compared with the classical calibration algorithm the proposed algorithm can quickly and accurately complete the calibration of the focal length of the camera.