基于先验模型的小波图象降噪方法

A Wavelet Denoising Method Based on a Priori Model

本文提出了一种基于小波系数先验模型的小波降噪技术。这种模型认为整个小波系数空间由服从零均值高斯分布的局部区域组成 ,具有空间自适应和局部化的性质 ,模型的参数可以利用实际观测数据 (包含噪声成分 )来估计。基于估计得到的小波系数先验模型可以设计小波域上的经验Wiener滤波器 ,并应用到图象降噪任务中。实验结果表明 ,此种方法可以取得比较好的降噪效果。

A wavelet denoising method based on a priori model is proposed in this paper. This model sees the whole wavelet coefficient space is composed of individual local areas, which obey the Gaussian distribution with zero mean. So this wavelet coefficient model is spatially adaptive and local. The parameters of this model are estimated based on the observed noisy data. Then the wavelet domain Wiener filter is designed and applied in image denoising. Experimental results show this method can get excellent effect.