Image retrieval based on color and texture

A novel image retrieval algorithm based on color and texture is presented in the paper. Firstly, the entropy of each color component is computed as color feature based on block truncation coding (BTC). Then, the traditional gray level cooccurrence matrix (GLCM) is extended to neighbor statistic moment level co-occurrence matrix (MLCM), and then the texture feature is extracted based on MLCM. Finally, the weighted color and texture feature are fused together for image retrieval. Experimental results show that the proposed method get higher performance than the other methods mentioned in the paper.