复合噪声图像去噪

Image Denoising for Mixed Noise

在现实的图像处理中,图像往往含有加法均匀或者高斯分布噪声的干扰.如果只使用邻域平均法或中值滤波法都不能很好的对图像进行去噪.提出了一种复合型去噪法.该算法首先利用分离像素法把不同噪声污染的像素区别开来,然后分别再用中值滤波或邻域平均法去噪.实验结果表明,此方法具有很强的可信度.

In fact,images usually are corrupted by Symmetrical noise or Guass noise, so good filtering effect can not be obtained if we only use average filter or median filter. In this paper,a new mixture filter method is proposed. The method classfies the pixels into two classes, then uses average filter or median filter to denoise. The test proved that the method is credible.