一种基于噪声概念的新的去噪方法  

A New Noise Removed Method Based on Noise Concept

作　　者：孔涛[1] 张亶[1] 

机构地区：[1]浙江大学数学系,杭州310027

出　　处：《工程图学学报》2004年第3期65-70,共6页Journal of Engineering Graphics

基　　金：国家自然科学基金资助项目(60202002)

摘　　要：在几乎所有的图像处理领域,在正式对图像处理之前,为了改善图像的质量,一般都要去除在图像传输等过程中产生的噪声。笔者介绍了一种新的去噪方法,该方法是基于噪声的概念提出的,对去除诸如椒盐噪声、随机噪声等效果都很好,尤其对随机噪声更显出其优势。该算法简单有效,容易实现。最后给出了在不同噪声量的情况下试验的对比图,经过大量的试验证明,这种新的算法能得到很好的去噪效果。In all most digital image processings fields, we always get rid of the noise by the digital images transmission in order to improve the quality of the digital image before we process the digital image formally. In this paper, a new noise removed method based on noise concept is introduced, and it demonstrates very good effect on removing noise, such as salt&pepper noise , stochastic pulse noise and so on, especially stochastic pulse noise .The arithmetic is simple and effective, and is realized easily. At last, we give a lot of the results of experiments in different noise case . By lots of experiments, we all can get good effects .

关 键 词：计算机应用 数字图像处理 概率统计 中值滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]