Study on method for image retrieval based on multi-feature fusion
Zhu Ru-sheng · Journal of Harbin University of Commerce · 2010
Single-feature-based image retrieval system has a lower performance.In this paper,multi-feature fusion based image retrieval is proposed.Through the non-interval quantization of HSV color space,extracts one-dimensional color feature vector and its cumulative histogram.Based on gray level co-occurrence matrix(GLCM) extracted the texture feature.Combination of two types of characteristics is done;an image retrieval based on multi-feature fusion is designed.This paper uses the method of similarity measurement for test of the color features and the texture features,as well as the multi-features.Experiment results show that multi-feature fusion method based on the color and texture features has a better performance.