针对椒盐噪声的开关模糊滤波器

自适应中值滤波随着椒盐噪声密度增加滤波图像细节损失较大。在开关滤波和自适应中值滤波的基础上提出了开关模糊滤波（SF）。SF用Max-Min算子检测噪声点，然后根据滤波窗口中正常点数量以均值方法或者T-S模糊方法去噪。仿真实验表明，开关模糊滤波比自适应中值滤波能更好地保护边界和细节。开关模糊滤波在消除噪声和细节保护之间取得了良好的平衡。

Switching based fuzzy filter for salt-and-pepper noise

Adaptive median filter(AM) was less efficient in detail-preserving when noise ratio is high.On the basis of standard median filter(SM) and AM this paper introduced a switching based fuzzy filter(SF).Noise pixels were identified by Max-Min operator,then according to the number of normal pixels,mean method or T-S fuzzy method was adopted to eliminate noise pixels.Experimental results show that this filter outperforms adaptive median filter in detail-protection and image smoothing.SF keeps a tradeoff between noise attenuation and detail-protection.