Research on feature matching algorithm based on three-dimensional information acquisition

Abstract

Image feature matching in three-dimensional information acquisition technology is an important aspect of machine vision, in order to achieve the real-time and matching robustness requirement, an image registration method based on SURF algorithm is proposed. First construct Hessian matrix based on box-type filter, use SURF algorithm to calculate the information of feature points, then decompose characteristics of feature points and match feature points. Experimental results show that the matching speed of SURF algorithm is 4 times as fast as the SIFT algorithm, greatly improves the real-time and robustness.