噪声图象最小方差模块滤波方法  

Noisy imag filtering algorithm based on block and minimum variance

作　　者：张晓明[1] 蒋大真[1] 卢宋林[1] 

机构地区：[1]中国科学院上海原子核研究所,上海交通大学

出　　处：《核技术》1996年第2期93-96,共4页Nuclear Techniques

摘　　要：用噪声图象平滑滤波的算法对一幅加了多种噪声的图象作了平滑。结果表明，本滤波算法在既不模糊图象边缘，又不损害图象内部细节的前提下，达到了很好的去除噪声、增强图象边缘的目的．In this paper proposed was a new filtering algorithm for noisy image.The most homogeneous neighborhood area around each point in an image was looked for using the minimum variance technique,and then each point was given the average.gray level of the selected neighborhood area.An experiment upon a noisy image with many kinds of noise was carried out.The result demonstrated that noises were removed well and the image edges were enhanced without blueing sharp edges nor destroying the details of a region.

关 键 词：噪声图象 图象滤波 最小方差 最大似然邻域 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]