基于阈值判断的自适应中值滤波算法

An improved adaptive median filter based on noises detection

针对标准的中值滤波算法在去除噪声与保留图像细节方面难以取舍的缺陷,在自适应中值滤波算法的基础上提出了一种改进的基于噪声点检测的自适应中值滤波算法.该算法在进行噪声点检测时采用了一种阈值判断法,充分利用了当前像素点与邻域像素点的灰度值之间的关系.结果表明,在噪声浓度较高时仍然可以区分噪声点与边缘点,滤波的同时有效地保护了图像的细节.

The standard median filter can not do well both in noise suppression and detail preservation,an improved adaptive median filter based on threshold detection was proposed.The algorithm took full advantage of the relationship between the center pixel and its neighbor pixels＇ gray value,and it set a threshold based on the difference between the center pixel and its neighbor pixels when detecting the corrupted pixels.Experiment results show that the new algorithm can distinguish the pixel corrupted or not,and it is both effective in noises suppression and detail preservation.