基于PCNN的脉冲噪声滤波算法

Pulse Noise Filter Algorithm Based on PCNN

鉴于噪声点和边缘点具有相似性,传统中值滤波、均值滤波很难对其进行区分,容易造成图像细节丢失。本文通过分析总结脉冲噪声的特点及脉冲耦合神经网络(PCNN)的工作机理,提出了一种基于PCNN的脉冲噪声滤波算法。首先利用PC-NN的脉冲传播特性检测出原始图像的噪声点和边缘点,然后利用噪声点和边缘点不同特点对其进行判断区分,若为噪声点进行中值滤波,边缘点则不做处理。实验结果表明该方法不但能有效的去除图像中的脉冲噪声,而且能很好的保护图像细节信息且提高了去噪后图像的峰值信噪比。

In view of the similarity of noise points and edge points,it is difficult for traditional median filter or mean filter to distinguish them,so some of image detail are lost.By analyzing and summarizing the characteristics of impulse noise and the the working mechanism of pulse coupled neural network（PCNN）,a PCNN-based filtering algorithm for impulse noise is proposed.Firstly,the noise points and edge points of the original image are detected by using the pulse propagation characteristics of PCNN,then distinguished by their different characteristics.After that,the noise points are dealt with median filter,but the edge points are not treated.Experimental results show that the method not only can effectively remove the impulse noise,but also can well protect the details and improve PSNR of the denoised image.