一种改进的方向加权中值滤波算法

An Improved Direction Weighted Median Filter Algorithm

为了去除图像中的随机值噪声,提出一种改进的中值滤波算法。基于图像每个像素的截断阈值,判别其是否为噪声像素,将判别结果用于修改方向加权中值滤波算法(DWM)的滤波器权重。仿真实验结果表明,对于去除图像中密度大于40%的随机值脉冲噪声,该算法比DWM算法具有更高的峰值信噪比,去噪效果较好。

In order to remove random noise from images,an improved median filtering algorithm is proposed.Based on the truncation threshold of each image pixel to determine whether it is a noisy pixel,and the result is used to modify the median filter weight of the DWM(directional weighted median)algorithm.The simulation results show that the proposed algorithm has a better denoising effect and a higher PSNR than DWM algorithm for removing random impulse noise with density greater than 40%in the image.