Two-step local Wiener filter using dual-tree complex wavelet transform
Xiaobo Zhang, Xiangchu Feng · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2011
Amongst various image denoising methods, the Wiener filter is known to be a mean-square optimal linear estimator for the true underlying image. The idea behind the Wiener filter frequently emerged in other image denoising algorithms. In this paper, we present a new denoising scheme for the Wiener filter. The algorithm is carried out by two subsequent steps. The output for the first step is used as the input for the second step. The ratio between the two step denoising amounts plays an important role for good performance. The relationship between the proposed method and the heat diffusion equation is established. The experiments show the proposed method outforms some of the denoising algorithm published currently.