基于非下采样Contourlet变换与TV模型混合图像去噪算法

在非下采样Contourlet变换的基础上,综合考虑全变差扩散和正态逆高斯模型,提出一种新的图像去噪算法.首先,对图像进行非下采样Contourlet变换,得到高频子带和低频子带系数.然后,对低频子带进行全变差扩散处理,对于方向带通子带,先通过分类准则对其进行分类,将其分为重要系数和不重要系数,对重要系数采样正态逆高斯建模,不重要系数采用高斯分布模型建模.实验结果证明,本文方法在视觉效果、峰值信噪比以及平均结构性上均优于许多算法.

A New Hybrid Image Denoising Algorithm Based on Nonsubsampled Contourlet Transform and TV Model

A novel image denoising method, based on nonsubsampled Contourlet transform, TV model and normal inverse Gaussian model is given. Firstly, the noisy image is decomposed into a set of multiscale and multidirectional frequency sub bands by NSCT, then applies TV diffusion to low frequency, according to classification standard to each high frequency sub band coefficients to important coefficients and unimportant coefficients. For important coefficients by normal inverse Gaussian model and unimportant coefficients by Gaussian distribution. Experimental results show that out algorithm better than the other algorithms in visual quality , peak signal--to--noise ratio and mean structural similarity.