Image denoising based on correlations of inter-scale coefficients in contourlet domain

A new image denoising algorithm based on contourlet transform is presented in this paper. The new approach takes the correlations of inter-scale contourlet coefficients into account in the process of shrinkage, and assumes that the noise-free contourlet coefficients are correlated to their parent coefficients which locate at a different scale. By computing the relativity coefficients across scales, we consider those with smaller values are more likely the noisy coefficients. And then we remove those coefficients whose magnitudes and corresponding relativity coefficients are both small. Experimental results demonstrate that the proposed algorithm can not only maintain the edges of an image, but obtain higher PSNR and better visual quality.