Spectral graph based image denoising methods

Some of the spectral graph based image denoising methods are reviewed and a Wiener filtering scheme in graph Fourier domain is proposed for improving image denoising performance achieved by these methods. The proposed Wiener fi lter is estimated by using graph Fourier coeffi cients of the noisy image after they are processed for denoising, to further improve the already achieved denoising accuracy as a post-processing step. It can be estimated from and applied to the entire image, or can be used patchwise in a locally adaptive manner. Our results indicate that the proposed step yields consistent accuracy improvement for di fferent choices of weighted adjacency and graph Laplacian matrices used in computing the graph Fourier transform and for diff erent processing methods used to denoise obtained transform coeffi cients. We obtain higher peak signal-to-noise ratio (PSNR) values than a state-of-the-art denoising method, known as the BM3D method, for some standard images.