Three-Dimensional Reconstruction Of The Stereo Matching Technology Research

A branch of the three-dimensional shape reconstruction of objects belonging to the computer vision technology, which the computer graphics, image processing and computer vision are combined to become a new research direction. It has a wide range of applications, both in the visual field of robotics, automation, virtual reality, CAD, and medicine, etc. can use it. How to get fast and accurate three-dimensional information of the object surface, and then the three-dimensional reconstruction based on this information, the recovered surface information of the object, is an important and very difficult research topic.With the continuous development of machine vision technology, stereoscopic vision has been more widely used, especially its structure binocular stereo vision is simple, easy to use, and many other advantages have been successfully used in industrial inspection, object recognition, workpiece positioning, robotics self-guided, aerospace and other fields. In many machine vision problem to be solved, the stereo matching problem is the most important and difficult. Any three-dimensional computer vision system includes a matching algorithm as its core, and therefore the study of the matching algorithm is important. Although research on the stereo matching problem for many years, but there are still all kinds of matching algorithm parallax fuzzy boundaries, shelter areas and weak texture region matching difficulties insurmountable problems.This paper introduces the basic concepts of computer vision and three-dimensional reconstruction, analyzes the research status of three-dimensional reconstruction techniques in various fields television, medicine, and other visual publications, research based on the reconstruction of three-dimensional vision of key technologies, do a set of three-dimensional reconstruction of the experiment, from image acquisition, feature extraction, matching stereo camera calibration to rebuild point. Article focuses on one of the key technologies that stereo matching algorithm is studied and simulation. Finally, a summary of all the work done in this paper, and pointed out the issues that need further research.