Image denoising algorithm based on edge-preserving self-snake model and wavelet-based PDE

Abstract

In this paper, a so-called edge-preserving self-snake model (EPSSM) that is able to remove noise while preserving edge features will be built, and using wavelet and partial differential equation (PDE), an integrated algorithm of wavelet-based PDE (IAWP) of image denoising is proposed. In integrated algorithm, The EPSSM is firstly used to remove noise of an input image, then we decompose the processed image by wavelet transform and its three high frequency coefficients are filtered by the EPSSM, finally, denoised image is reconstructed using inverse wavelet transform. The denoising performance of two proposed algorithms is measured according to PSNR values, and the experiments show that our methods have a better performance than others. © 2013 Springer-Verlag.