An easy dominant color extraction and edge valley histogram for image retrieval

This paper adopts a content-based image retrieval method based on an efficient combination of dominant colors and edge valley histogram. The dominant colors, which are the most frequently appeared colors in RGB color space, are easy to extract and closed human perception. Edge valley histogram represents the motifs of the edge and valley blocks in images. The two histograms with the discriminative motifs are easy to compute. Experimental results show that the proposed method has higher retrieval accuracy than Lu and Chang´s method, BDIP and BVLC, MCM and EDH even though its number of features is not greater than theirs. So the features can represent images discriminatively.