176 Four Features Contourlet Texture Retrieval

Contourlet transform is a non-tensor product special transform and has better performance in directional information representation than wavelet and superior to wavelet at retrieval application fields. In order to improve the retrieval rate further, we will propose a new contourlet transform based texture image retrieval system in this paper. In the system, standard deviation, energy, skewness and kurtosis of each subband coefficients in contourlet domain are cascaded to form feature vectors, and the similarity metric is Canberra distance. Experimental on a 109 texture image database from MIT results show that this contourlet transform based image retrieval system is superior to that of the contourlet transform with energy and standard deviations features widely used today under the same system structure.