摘要
针对脉冲噪声的特征,提出结合检测技术与四阶偏微分方程的MSDLC算法.该算法采用"先检测再复原"的两步策略:检测阶段,利用统计的思想定位受到脉冲噪声污染的噪声像素;复原阶段,采用LC四阶偏微分方程对噪声像素点进行复原.采用的LC四阶偏微分方程根据方程性质使用适当的系数函数.实验表明,MSDLC算法可以在脉冲噪声去除方面和保持图像细节方面获得极好的平衡.
According to the characteristics of the impulse noise,proposes a new algorithm called MSDLC Algorithm,which combined with detective tenology and fourth-order Partial Different Equation（PDE）.The new algorithm takes the measure of ＂detection firstly then restoration＂.In the detection phase,MSDLC algorithm takes advantage of statistic idea to locate the pixels which are polluted by the impulse noise.In the restoration phase,MSDLC algorithm makes use of LC fourth-order PDE to remove the noises,and the chosen LC fourth-order PDE uses appropriate coefficient function according to the property of LC fourth-order PDE.The experiments show that MSDLC algorithm can achieve a good trade-off between noise removal and edge preservation.
出处
《闽江学院学报》
2009年第5期53-57,共5页
Journal of Minjiang University
