Abstract: In order to improve the quality of underground images acquisition, according to the advantages and disadvantages of the current median filter algorithm, a new denoising algorithm is proposed which eliminates salt and pepper noise from underground images that are corrupted by salt and pepper noise. The new algorithm first adjusts noise points and non-noise points in image, then determines the size of filtering window adaptively on the basis of the number of noise points in window, and gives corresponding weight to each pixel points in the filtering window adaptively by certain rules, finally makes full use of weighted median filtering algorithm processing the noise points. Computer experimental results clearly show that the proposed algorithm outperforms many of the existing median filter algorithm in terms of effectively eliminating the noise and protecting image detail well.

Key words: median filter, salt and pepper noise, adaptive, detail preserving

摘要： 为了提高井下图像采集的质量，针对目前改进中值滤波算法的优缺点，提出了一种新的去除井下图像椒盐噪声的算法。该算法首先判断出图像中的噪声点和非噪声点，然后根据窗口内噪声点的密度大小自适应地确定滤波窗口的大小，并按照一定的规律赋予窗口内像素点不同的权重，最后采用加权中值方法处理图像中的噪声点。计算机模拟实验证明该方法不仅能有效地去除不同密度的椒盐噪声，而且能很好地保持图像的细节，滤波效果比已提出的中值滤波算法更好。

关键词: 中值滤波, 椒盐噪声, 自适应, 细节保护