摘要
针对受脉冲噪声污染大(大于50%)的图像,提出运用多幅序列图像的点对点噪声检测算法.首先利用MMEM算法判断噪声点与非噪声点,再把非噪声点拷贝到输出图像,通过实验得出了噪声密度与所需要图像幅数之间的关系.通过噪声密度判别公式的引入,实现对噪声图像的自适应处理,实验显示该方法优于传统滤波算法.
Focused on the highly corrupted （〉50%）images by impulse noise,a noise detective algorithm pixel by pixel is proposed using sequential images . Firstly, the good pixels and noise pixels are judged with MMEM algorithm, and the good pixels are copied to the output image. Then, caculating the relationship between the noise density and the number of the images needed, by importing the estimating operator of noise density, an adaptive algorithm for image denoising is proposed. The experimental results show that the method is better than traditional algorithms.
出处
《江南大学学报（自然科学版）》
CAS
2007年第5期532-536,共5页
Joural of Jiangnan University (Natural Science Edition)
基金
国家自然科学基金项目(60672072)
浙江省自然科学基金项目(Y106505)
关键词
脉冲噪声
高污染
序列图像
中值滤波
impulse noise
highly corrupted
sequential images
median filter
