Color Image Retrieval Based on Visual Saliency Map
HU Sheng-xiong, Shang Zhao-wei, Taiping Zhang, Weibin Yang · Jisuanji gongcheng · 2012
【Abstract 】In order to improve efficiency of image retrieval, a retrieval method based on visual saliency map is proposed in this paper. The background information unrelated to retrieval task in original image is removed according to the visual saliency map, and remains the region of interest. The BDIP-BVLC moments in the wavelet domain is used to extract the image features. The secondary inquires measure strategy is adopted as the distance function. Experimental result shows that the proposed method outperforms the method based on saliency weighted. It has higher retrieval accuracy.