Image Retrieval Algorithm Based on Color Block Histogram

Abstract

To further enhance the efficiency of color histogram - based image retrieval algorithm, this paper presents a new algorithm which looks for maximum color lump, extracts its local histogram, and integrates with the original histogram. The algorithm adds spatial distribution characteristics of color while retaining the original algorithm, thus reducing absolute dependence on the color for retrieval results. The simulation results show that this algorithm has better precision and recall and has especially remarkable effect for image retrieval of apparent color lump.