基于改进PCNN模型的椒盐噪声级化滤波方法
A graded filtering method based on improved PCNN model for salt and pepper noise
-
摘要: 为有效滤除椒盐噪声同时保留图像的灰度细节,提出了一种椒盐噪声滤波算法.首先利用改进PCNN简化模型进行一次点火过程,定位灰度图像中被噪声污染和未被噪声污染的像素点,然后采用提出的级化中值滤波算法对已定位的噪声点进行滤波而保持其他像素点不变.仿真实验表明,提出的方法对不同强度的椒盐噪声图像均有较好的滤波性能.比较已有的滤波算法,该算法能在高噪声强度时有效滤除噪声并同时很好地保留图像的边缘细节,实验结果证明了算法的有效性.Abstract: An efficient image filter algorithm was proposed to filter the salt-and-pepper noise and keep the detail of gray scale image,simultaneously.At first,an improved and simplified PCNN model was used to locate the noisy pixel by a firing process without circulation.Then the graded median filter algorithm was proposed to filtering the noisy pixel while keep the non-noise pixel untouched.Computer simulations indicated the proposed method demonstrated the good filtering performance when filter different intensity of salt-and-pepper noise image and proved the validity of the algorithm by objective and subjective valuation.The proposed method showed better filtering results than existing methods when the noise intensity was high.
下载: