In the variational denoising model,the total variation (TV) model can preserve the edges of image,but it can lead to staircase effect.High-order model can effectively overcome the staircase effect and can keep the contrast ratio and the angle point of image.But the diffusion of model is not uniform enough in the smooth region.In this paper,the TV and TC (total curvature) model are combined to construct a new model which has good performance in preserving edge,angle point and contrast ratio and can overcome staircase effect of TV model.In this paper,we designed a fast Split Bregman algorithm by introducing auxiliary variables and Bregman iteration parameters.The experimental results show that the new model's efficiency is better than the TC model's.And the model's convergence iterations are less than TV and TC model's.The model has good performance in the edge,comer and smooth degree.Signal-to-noise ratio of new model is better than tradition models' results'.%在变分去噪模型中,TV(total variation)模型对于图像的边缘具有较好的保持效果,但处理后的图像会有阶梯效应.高阶模型可以有效地克服阶梯效应,并且可以保持对比度和角点,但在光滑区域扩散不够均匀.现将TV和TC(total curvature)模型相结合,从而构造出新的模型,同时具有较好的边缘、角点和对比度的保持能力,克服TV模型的阶梯效应,通过引入辅助变量和Bregman迭代参数设计了快速Split Bregman算法.实验结果表明,所提出的新模型计算效率要优于TC模型,迭代收敛步数要小于TV和TC模型,在边缘、角点以及光滑度等特性方面取得了良好特性.相对于上述传统模型,新模型处理后图像的信噪比均有提升.
展开▼
