Improvements on Sparse Coding Shrinkage and Contourlet Transform for Image Denosing

In this work, according to the disadvantages of sparse coding shrinkage and contourlet transform, we investigated the use of sparse coding shrinkage in conjunction with contourlet transform for denoising image data and introduced a new image denoising algorithm. The new algorithm based on linear noise model and excellently solves the denoising of image that contains additive noise with unknown variance. Experimental results show that this new algorithm is indeed effective and efficient. Compared with other denoising methods, the algorithm is much better for it enhances the value of SNR, reduces the value of MSE, and obtains a better quality of image reconstruction.