Image denoising with anisotropic bivariate shrinkage
Signal Processing
Hi-index | 0.00 |
In the recent years, denoising based on the spatially adaptive algorithms that employ anisotropic adaption have been developed. These methods are able to match to the local statistics, preserve the edges and truly remove the noise from the texture of the images. On the other hand, a huge proportion of image enhancement methods are implemented in the sparse domains (e.g., wavelets, curvelets, contourlets and steerable pyramid decomposition) due to impressive properties of these transforms such as heavy-tailed nature of marginal distribution, locality and multiresolution. In this paper we try to establish a relation between two mentioned approaches by estimating the local variances of steerable pyramid coefficients using a shape-adaptive window.