为解决模糊图像复原技术经典算法受噪声等高频成分影响较大的问题，文中提出一种简单快速处理含特殊噪声模糊图像的方法。当噪声叠加到图像中，图像赖以估计模糊核参数的频域特征减弱甚至消失。一般去除噪声的方法在滤除噪声的同时也除去了一些可用于模糊参数估计的重要信息，因此，含噪图像通常会寻求它法处理。但文中研究发现，对于某种特殊的噪声例如椒盐噪声，使用针对于它的滤波方法，例如中值滤波，信息损失量极小，因此含特殊噪声图像可以按此方法处理。结合有针对性的滤波方法去除特殊噪声，减少信息损失，从而快速恢复图像。仿真实验结果表明，对于含噪图像，该算法比以往算法更优越。%In order to solve the problem that classic algorithms of fuzzy image restoration technology are greatly influenced by noise,pro-pose a simple and rapid processing method for fuzzy image which contains special noise in this paper. When the noise is added in the im-age,the frequency domain feature of image estimating fuzzy kernel parameter weakens,even disappears. The general noise removal meth-ods can remove some important information for fuzzy parameter while filtering the noise,therefore,noisy image often find other method. But in this paper,find that for some special noise such as salt and pepper noise,use special filtering method such as median filtering,the information loss is very few,therefore,image contains special noise can be treated by this method. Combined with targeted filter method to remove noise,reduce the information loss and restore image. Experimental results indicate that the algorithm for noisy image is superior to the former algorithm.
展开▼
