Image denoising in the wavelet domain using Improved Neigh-shrink

Denoising of images corrupted by Gaussian noise using wavelet transform is of great concern in the past two decades. In wavelet denoising method, detail wavelet coefficients of noisy image are thresholded using a specific thresholding function by comparing to a specific threshold value, and then applying inverse wavelet transform, results in denoised image. Recently, an effective image denoising method has been proposed called Neigh-shrink that exploits the interscale dependency of wavelet coefficients. In this paper, we extend Neigh-shrink denoising method by proposing a new thresholding scheme. Experimental results show that our method outperforms classical Neigh-shrink visually and in the terms of PSNR.