基于模糊方法的随机值脉冲噪声去除算法

Random Value Impulse Denoising Agorithm based on Fuzzy Method

随机值脉冲噪声脉冲强度的动态范围为整个灰度空间,通常为均匀随机分布,因此通常不知道它的先验分布。一般算法很难区分对比度差的信号与噪声。本文针对随机值脉冲噪声的特征,构建了噪声描述描述模型,利用模糊系统理论描述随机值脉冲噪声的存在特征,通过像素多邻域、自适应窗口检测噪声,运用最小化价值函数滤除噪声,实验证明本文提出的算法是一种有效的随机脉冲噪声检测与去除方法。

The intense of random value impulse noise is covered whole grayscale space,which obeys the uniform distribution. The general algorithm is difficult to distinguish contrast difference between signal and noise. In this paper,a noise model was built,which utilized the fuzzy systems theory to describe noise characteristics. Noise was detected by using pixel multiple neighbor and adaptive windows. Finally,the denoising was done with minimizing the cost function. Experiments show that the proposed algorithm is an effective methods for removing the random value impulse noise.