Study on method for image retrieval based on feature fusion

Single-feature-based image retrieval system of content-based video retrieval system has a lower performance. In this paper, a algorithm of image retrieval based on NMI invariable feature of dominant color and texture feature is proposed. The shape information and the spatial distribution of dominant color can be described by the NMI invariable feature of dominant color so the content feature of image can be attained easily and effectively. In order to enhance retrieval efficiency use improved gray level co-occurrence matrix(GLCM) extracted the texture feature. Combination of two types of characteristics is done. And this paper uses the method of similarity measurement for the multi-features. Experiment results show that the algorithm has a better performance.