Image Stitching Based on Binocular Vision

Abstract In the field of photography, image stitching is used to generate panorama. In most cases, the position and intrinsic parameters of the camera can not be obtained, so the transform matrix can only be calculated using the information of the image. In industrial machine vision detection system, the information of the camera can be obtained by calibration. It is feasible to improve the speed of image stitching algorithm by using the information of camera calibration. An image stitching algorithm based on binocular vision was proposed. Compared with the traditional stitching algorithm based on image information only, it greatly improved the running speed of the algorithm. The effectiveness of the algorithm was verified by experiments.