Image Retrieval Based on Weighted Block Color Histogram and Texton Co-occurrece Matrix

Abstract

Keywords: image retrieval; HSV color histogram; texton co-occurrence matrix; feature extraction. Abstract. Image retrieval is an important topic in the fields of image processing, pattern recognition and computer vision. In this paper, we proposed a new and efficient method of image retreival by combining the color and texture features. We used HSV color histogram for color feature extraction and texton co-occurrence matrix for texture feature extraction. In color feature extraction, the image was divided into nine unequal blocks first and then for each block HSV color histogram was computed. In texture feature extraction, the gray level image was filtered with four predefined textons first and then we used co-occurrence matrix to extract the texture feature of the filtered texton image. After feature extraction, we used Euclidean distance to measure the similarity of the query image and images in the database. We emploied Corel 1000 database for our experiment, the experimental results show that our proposed method performs better than the previous methods.