首页 | 本学科首页   官方微博 | 高级检索  
     

基于中值滤波与灰度级形态学的综合去除椒盐噪声方法
引用本文:岳俊华,李岩,纪玉波,郭立红.基于中值滤波与灰度级形态学的综合去除椒盐噪声方法[J].长春理工大学学报,2004,27(3):70-73.
作者姓名:岳俊华  李岩  纪玉波  郭立红
作者单位:辽宁石油化工大学信息工程学院,抚顺,113001;中科院长春光学精密机械与物理研究所,长春,130022
基金项目:国家 8 63计划 ( 863 -70 3主题基金 )
摘    要:为了消除图像上的椒盐噪声,并尽量保持图像的清晰度,提出了一种在中值滤波基础上,经过灰度直方图的分析,再采用合适的灰度级形态学运算去除图像中的椒盐噪声的新方法。该方法,既弥补了中值滤波在椒盐噪声概率高的情况下滤噪能力下降快的缺点,又改善了灰度级形态学在椒盐噪声概率高的时候,达不到有效滤噪的不足。实验表明,该方法滤除椒盐噪声效果优于单独使用中值滤波或形态学滤噪,在椒盐噪声概率超过0.2时,这种优势尤为明显。

关 键 词:椒盐噪声  中值滤波  直方图  灰度级形态学
文章编号:1004-485X(2004)03-0070-04
修稿时间:2004年6月18日

An Method for Removing Image Salt & Pepper Noise Based on Median Filter and Gray Morphology
YUE Junhua,LI Yan,JI Yubo,GUO Lihong.An Method for Removing Image Salt & Pepper Noise Based on Median Filter and Gray Morphology[J].Journal of Changchun University of Science and Technology,2004,27(3):70-73.
Authors:YUE Junhua  LI Yan  JI Yubo  GUO Lihong
Affiliation:YUE Junhua1 LI Yan2 JI Yubo1 GUO Lihong2
Abstract:Information are playing more and more important role i n our life.Most of them are images.However, salt & pepper noise occur during ima ge generation, communication and transformation, and affect image quality.By us ing median filter or gray morphology filter method, salt & pepper noise can be r emoved.But along with the increase of the noise, these two methods can not deal with efficiently.Where there is too much salt &pepper noise, the median filter may have a lower capability to remove it; meanwhile, morphology filter method m ay also be fuzzy to the image.In this paper, the new method is described.Firstly , processing the salt and pepper noise by median filter; secondly, analyzing the processed image by gray histogram; thirdly, using a suitable morphology means t o remove the noise. The experimental results show that this method can provide m uch better performance compared with the other methods. Especially, the merit i s more obvious when the noise probability exceeds 02.
Keywords:salt & pepper noise  median  filter  gray  histogram  gray  morphology
本文献已被 CNKI 维普 万方数据 等数据库收录!
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