摘要
在图像去噪处理算法中,一般的方法对于边缘信息总是存在不同程度的信息丢失。本文考虑了一种新的去噪处理的方法——保信去噪。该方法在均值去噪基础上,对其所求的均值求导,判断其均差来确定该点的像素值。此方法有效的克服了均值去噪中图像的边缘模糊的缺点,具有较好的鲁棒性。试验结果表明:该方法的效果明显。
Regarding the edge information, general methods always result in some degree information loss in the process of image denoising. This paper introduces a new denoising method of information preserving and denoising. Based on the mean denoising, this denoising method judges its variance and mean＇s difference to determine pixel value, it effectively overcomes the picture edge fuzzy shortcoming of the mean denoising and has a better robustness.
出处
《红外技术》
CSCD
北大核心
2006年第7期415-418,共4页
Infrared Technology
基金
成都理工大学青年基金项目(编号:2005qi24)
关键词
保信去噪
均值去噪
图像处理
information-preserving and denoising
mean denoising
image processing
