A Novel Gradient L0-Norm Regularization Image Restoration Method Based on Non-local Total Variation

Abstract

This paper proposes a novel image restoration method based on non-local total variation (TV). Firstly, the image is divided into two types of regions by the gradient L0 norm. The one regularized by the local TV term contains edges and flat regions, the other regularized by the non-local TV term contains rich image details. Then, in order to simplify complex numerical algorithms, we adopt the alternating direction method of multipliers (ADMM) algorithm to optimize the object function. Finally, we carry out comparative experiments with several recent state-of-the-art methods to verify the performance of the proposed method. Experimental results show that the proposed method has better performance in the efficiency and get a good balance between balance between easing staircase effects and retaining image details.