基于小波域马尔可夫先验模型的图像去噪方法

Image Denoising Based on Wavelet Domain Markov Random Field Prior Model

提出了一种基于各向异性马尔可夫随机场(Markovrandomfield,MRF)先验概率模型的图像去噪方法。该方法利用图像小波子带的方向性特点以及小波系数尺度内和尺度间的相关性,将小波系数的分布特征建模为一种各向异性MRF先验概率模型。通过在贝叶斯框架中采用这种先验概率模型可以得到一种具有空间自适应性的贝叶斯萎缩函数。利用这种萎缩函数可以实现对小波系数的修正。实验结果表明利用该方法进行图像去噪能够取得良好的效果,同时可以有效地保留图像的细节。

An image denoising method was proposed based on an anisotropic Markov random field （MRF） prior model. This method modeled the configurations of the wavelet coefficients as an anisotropic MRF. This model took into account inter- and intrascale dependencies between wavelet coefficients and it was adaptive to the wavelet subbands corresponding to three orientations in the image. Based on this prior model in a Bayesian framework, a spatially adaptive Bayesian shrinkage function was obtained and each modified coefficient was decided separately. Experimental results demonstrate this method improves the denoising performance and preserves the details of the image.