摘要
针对自适应中值滤波算法的缺陷——对高密度椒盐噪声图像滤波后留下黑色斑块,提出了一种分阶段中值滤波算法.该算法对图像执行两次小窗口的滤波操作,相较于采用较大窗口的滤波,其在有效去除噪声的同时降低了结果图像的模糊程度.先对所有噪声点进行一次中值滤波消除了盐粒噪声,再用窗口内非噪声点的灰度中值代替胡椒噪声点的灰度值以去除黑色斑块.最后的仿真实验结果表明,本文算法既有像自适应中值算法一样滤除低密度椒盐噪声的良好性能,又有对高密度椒盐噪声图像的降噪能力.
A grading median filtering algorithm was proposed to offset the defect of adaptive median filter （ AMF） that it left some black plaque after filtering images corrupted by high density salt and pepper noises. Through twice filter to the noise image with small size win-dow,compared with bigger ones, it reduced the blur degree of result image. For the first time,it eliminated the salt noise by using median filter （ MF ） to noise pixels, and then wiped off the black plaque by replacing pepper noise pixels with the median of the noise free pixels in its 8-neighborhood. Lastly, the simulation result shows, our algorithm either has the good capability to filter the low density noises as well as AMF or has the ability to filter higher density salt and pepper noises in image.
出处
《南华大学学报（自然科学版）》
2013年第3期66-70,77,共6页
Journal of University of South China：Science and Technology
关键词
椒盐噪声
中值
分阶段去噪
salt and pepper noise
median
grading de-noising
