An Independent Component Analysis Algorithm Based on Block-wise Contourlet Transform
基于分块Contourlet变换的图像独立分量分析方法

Based on the characteristics of good sparsity and capturing effectively the smooth contours in natural images for Contourlet transform, a new independent component analysis algorithm is proposed by using the block-wise Contourlet transform in this paper. Experimental results show that the proposed algorithm is able to achieve a better performance in image separation.