Image Matching Based on Unification
In the image retrieval which is based on content, the two most important operations are extracting the stable image features and matching them. This paper adopts Scale Invariant Feature Transform algorithm (namely SIFT algorithm) to extract image feature points, and then using the theory of unification to perform reliable matching between different views. The features which are extracted from scale-invariant space are invariant to image scale and rotation, and are shown to provide robust matching across a substantial rang of affine distortion, addition of noise, and change in illumination. This paper introduces the unification into the features matching system, the result verify this algorithm has a well adaptability to various conditions, and it improves the match accuracy, at the same time reduces the amount of calculation.
Unification Scale-invariant space Scale Invariant Feature Transform Feature points
Xiaoli Li Xiaohong Wang Chunsheng Li
Department of Computer Sciences, Communication university of China, Beijing,100024,China
国际会议
昆明、丽江
英文
825-828
2011-04-15（万方平台首次上网日期，不代表论文的发表时间）