首页 | 本学科首页   官方微博 | 高级检索  
     

直方图加权均值滤波器
引用本文:杨群生,黄继武,康显桂,ZHOU Wang.直方图加权均值滤波器[J].电子学报,2004,32(7):1108-1111.
作者姓名:杨群生  黄继武  康显桂  ZHOU Wang
作者单位:1. 中山大学电子系,广州 510275;2. 华南师范大学计算机系,广州 510631;3. Lab for Computational Vision,New York University,New York 10003,USA
基金项目:国家自然科学基金,国家自然科学基金,国家重点实验室基金
摘    要:本文提出了一种适合于消除图像盐椒噪声的滤波器——直方图加权均值(HWM)滤波器.该算法以加权均值滤波器为基础,利用被污染图像的直方图函数作为权值进行加权运算.实验表明,对于噪声率在5%到90%的的噪声图像,HWM滤波器具有良好而稳健的去噪效果,当噪声率超过70%时,其优越性更加突出.

关 键 词:直方图  图像处理  直方图加权均值滤波器  
文章编号:0372-2112(2004)07-1108-04
收稿时间:2002-11-18

Histogram Weighted Mean Filter for Image Processing
ZHOU Wang.Histogram Weighted Mean Filter for Image Processing[J].Acta Electronica Sinica,2004,32(7):1108-1111.
Authors:ZHOU Wang
Affiliation:1. Dept.of Electronic and Communication Engineering,Sun Yat-Sen Univ,Guangzhou,Guangdong 510275,China;2. Dept.ofComputer Science,South China Normal Univ,Guangzhou,Guangdong 510631,China;3. Lab for Computational Vision,New York Univ,New York 10003,USA
Abstract:A new weighted mean filter,histogram weighted mean (HWM) filter,is proposed to restore images corrupted by salt pepper impulse noise.The idea is based on weighted mean filter,which uses the histogram function of the corrupted image as weight.Experimental results demonstrate that the proposed HWM filter is effective androbust for different noise ratio,ranging from 5% to 90%.When noise ratio exceeds70%,HWM filter gives superior performance compared with other known filters.
Keywords:histogram  image processing  histogram weighted mean filter  
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《电子学报》浏览原始摘要信息
点击此处可从《电子学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号