Content-based image retrieval based on wavelet-decomposition and feedback
Qunxiong Zhu · Journal of Beijing University of Chemical Technology · 2006
This paper proposed an image retrieval method based on wavelet-compress and feedback which can improve the performance of a content-based image retrieval system.The proposed method compresses images based on sub-band coding,and then exacts compressed images features such as color,texture and shape.Correction and validity can be improved by the feedback of clustering correlativity.Experiment results prove that although in shape feature space the proposed method has worse performance than the basic method based on original image,in color and texture feature space the proposed method has higher performance than the basic method.