An image denoising model using an adaptive total variation regularization
Abstract
In this paper, we present a denoising model based on the variation approach. To better denoise and preserve edges, we propose an adaptive total variation term based on a weighted function, values of which are adapted to the features of pixels. The mean curvature is used to describe the features of images and adjust values of the weighted function, i.e. strength of smoothing of the model. The Split Bregman is applied to solve the minimization problem. The numerical results demonstrate that the proposed model yields better denoising performance compared with the classical model in both terms of quantitative and qualitative criteria
Moraref