The General Variation Models of Additive and Multiplicative Noise Removal of Color Images and Their Split Bregman Algorithms
Zhenkuan Pan, Cuiping Wang, Weibo Wei, Chao Lu · 2011
The general variation diffusion models for additive and multiplicative noise removal of color images are proposed and their Split Bregman algorithms are designed via introducing auxiliary variables and Bregman iterative parameters, which lead to simple Poisson equations and analytical soft threshold formulas of the original minimization problems. The MTV (Multichannel Total Variation) and MPM (Multichannel Perona Malik) regularizations are considered as two examples of the proposed general regularizer and used for additive and multiplicative noise removal of color images with different kinds of noise. Finally, some numerical experiments are provided to validate the models and algorithms proposed in this paper.