一种基于小波域维纳滤波的图像复原算法

摘要

In the image restoration algorithm,a simple spatial filtering of frequency domain filtering algorithm is always easy to implement,but often requires more prior knowledge of image degradation.Iterative algorithm based on Bayesian theory is always effective in image restoration,however it is time-consuming.Spatial filtering or frequency methods are less timeconsuming,but the effect of restoration is not very good.Against this contradiction,for one of the characteristics of wavelet transform is multi-resolution,an image restoration algorithm was presented based on wavelet domain Wiener filtering using different algorithms to recovery different wavelet coefficients.The experimental results show that the method increase the efficiency of the recovery and ensures the quality of image restoration as well.Therefore,the algorithm is effective.%在图像复原算法中,单纯的空间域或者频域滤波算法简单易实现,但需要较多图像退化的先验知识.基于贝叶斯理论的迭代复原算法复原效果好,但耗时长.针对这一矛盾,利用小波变换的多分辨特性,对不同的小波系数特性采用不同的算法进行恢复,提出了一种基于小波域维纳滤波的图像复原算法.实验结果证明,所提方法在保证图像复原质量的同时相对提高了复原算法的效率,是一种有效的方法.