摘要

One of the most important problems in color Content-based Image Retrieval (CBIR) is the lack of an effective way to represent both the color and spatial information of an object. In this paper, we propose a material feature based algorithm for color image indexing and retrieval. Firstly, a physical parameter-track, which can be used to describe the character of a material is defined. Secondly, the equation of the track was gained by genetic programming. Thirdly, the material was distinguished by the material track (MT). In this way, we can retrieve a special material from an image database. Experimental results have shown that the algorithm is effective and efficient in color image indexing and retrieval