Comparisons of image matching methods in binocular vision system

Binocular vision system can be divided into different parts, and image matching is the one and very important component in this system. In this paper, the basic procedure of image matching algorithm is researched, and the classification of the image matching is analyzed due to the difference of matching primitives. Then the characters of the surveyed methods, including the advantage, disadvantage and application prospects are compared. Based on the analysis and experiment results, the algorithms based on Harris angular points, SIFT, SURF is compared. At lost, the next research direction and hotspot topic of image matching is discussed.