Research Of Character Detection And Fast Matching Of Three-Dimensional Reconstruction Base On Binocular Parallax

Feature detection and matching is important part of the three-dimensional reconstruction and the contents of this theme include: virtual reality, pattern recognition, image retrieval, video animation and many other important areas of research and application, is precisely because of this in recent years become the field of computer vision and computer graphics in one of the focus.The purpose of this theme is based on binocular vision system target scene reconstruction of three-dimensional structure, surface structure and the recovery model grid, and reconstruction results are applied to solve practical problems. The key technologies include: camera calibration parameters, image preprocessing, feature point extraction, feature matching, dense matching and 3D reconstruction. This paper uses image processing, computer graphics, computer vision, pattern recognition theories and a study.The main research topics are mainly three-dimensional reconstruction of the feature detection and feature matching, through analysis of classical Harris operator and classical matching method, made some improvement methods, including Harris operator adaptive detection method, based on binocular parallax of the fast matching method and to guide matching. And the experiment proved the feasibility of these improved algorithms and effectiveness.