An Efficient Implementation of Neighborhood based Wavelet Thresholding For Image Denoising

universal threshold value and identical neighboring window size in all wavelet subbands which may result in overly smooth images. To overcome NeighShrink weakness, Z. Dngwen, and C. Wengang[9] proposed SURENeighShrink that search for optimum threshold value and neighboring window size for every subband according to Stein's Unbiased Risk Estimate(SURE) method. It has been shown that the denoising performance of SURENeighShrink is considerably superior to NeighShrink and also outperforms SURE-LET[4] which is one of the best term-wise denoising algorithm that also based on SURE. The main drawback of the SURENeighShrink is its high computation cost when searching for optimum threshold value and neighboring window size for every wavelet subband. In this paper, we proposed an efficient implementation of SURENeighShrink that overcomes its computation complexity. The experimental results show that the proposed denoising algorithm achieves comparable denoising performance over the wide range of images and noise levels. Results also show that the proposed method outperforms NeighShrink considerably.