期刊名称：International Journal of Computer Science and Network
印刷版ISSN：2277-5420
出版年度：2017
卷号：6
期号：3
页码：320-324
出版社：IJCSN publisher
摘要：Denoising of images has been a successful research topic for various image processing applications. Image denoising is
basically restoration of images, where the unwanted noises causing degradations are removed to obtain a visually effectual high quality
image. The majority existing image denoising algorithms failed to focus on the diminishing edges whilst noise reduction. The net
effect is the low quality denoised image. This paper tackle the edge preserving problem by presenting SAIST (Spatially Adaptive
Iterative Singular-value Thresholding) image denoising algorithm incorporating bilateral filtering. In this work a two-fold approach is
adapted. First is preserving edges through bilateral filtering. A non- maximum suppression on the smoothed image and morphological
dilation to stretch the edges are performed. Second is image denoising using iterative regularization and singular valued decomposition
(SVD) for estimating signal variances. The pragmatic results and better computational efficiency do better than several state-of-theart
image denoising algorithms.
