Research on similarity measurement method of statistical sorting region based on differentiation and location information

Abstract

In the process of content-based image retrieval, making reasonable matching strategy and similarity measurement is the key technology to ensure the accuracy, quality and matching speed of image matching and retrieval. In this chapter, we focus on the problem of image feature selection and similarity measurement in content-based image retrieval technology, and propose a regional similarity measurement method based on differentiation and location information by using large image plane position constraint. This method mainly includes the following two aspects. One is the feature selection method based on global feature differentiation; the other is the similarity measurement method based on location information. Finally, we use the idea of statistical sorting to synthesize the image matching strategy and similarity measurement criteria through the set of feature points in the hot area. Thus, the problem of image similarity measurement from single feature point matching to target region can be solved.