Stone images retrieval based on color histogram

Stone images´ color features are chosen for its retrieval and indexing in this paper. Cross color histogram, annular color histogram, and angular color histogram are respectively combined with HSV color space to accord with human´s visual uniformity. In HSV color space, H, S, V three components carry on unequal interval quantization to improve retrieval precision and efficiency, and then different weight values are given and adjusted to different color channel. Stone images are sorted according to their similarity in retrieval results. Experiments showed that cross color histogram in HSV color space has excellent performance on stone images retrieval.