Research on 3D Reconstruction in Binocular Stereo Vision Based on Feature Point Matching Method

Abstract

Computer vision as an important branch of computer science and artificial intelligence has made rapid progress in the past thirty years. Binocular stereo vision is one of the most important parts in computer vision. Binocular stereo vision can well simulate human eyes stereoscopic perception of three-dimensional objects, and has been widely used in various fields of industrial automation production and practical life. Binocular stereo vision technology is a comprehensive technology, whose knowledge includes optics, physics, image processing, computer, artificial intelligence and electronic technology, and other fields of content. Camera calibration and feature point matching are difficult and key points in binocular stereo vision technology. In this paper, the 3D reconstruction of binocular stereo vision based on feature point matching method is discussed. The main research includes the following contents: binocular stereo vision system principle, feature matching algorithm research and 3D reconstruction system implementation. This project puts forward a set of feasible algorithms and finally can get a good three-dimensional reconstruction effect based on the analysis and research of a series of algorithms.