基于各向异性的二维云模型滤波

针对云模型滤波器在去除椒盐噪声的过程中没有考虑到图像各向异性的缺点, 提出了一种新的图像去噪方法。该方法首先利用云模型滤波器的噪声检测器对噪声图像进行检测, 然后对认定为噪声的图像部分利用基于二维云模型的模糊加权滤波算法, 并结合噪声周围非噪声像素的空间位置信息和像素值信息进行图像复原。将这种新的算法与原始方法进行相同环境下的对比实验。结果表明, 对于椒盐噪声密度为10%～50%的测试图像, 虽然在计算效率方面该算法略逊于原始算法, 然而在处理的过程中, 其展现了图像各向异性, 同时也更好地保持了图像的细节信息; 而且图像细节越复杂, 提升效果越明显。

Anisotropy-based filter with two dimensional cloud model

In order to overcome the neglect of the cloud model filter in image anisotropy, this paper presented a novel image denoising filter. The proposed filter used a noise detector based on cloud model to identify the noise first, and then it restored the noise image by using a two dimensional cloud model filter with combining the gray values and the location information of pixels. The experimental results show that, compare with the cloud model filter, although the proposed filter is slightly lower in computational efficiency, however, it shows image anisotropy in the processing, and preserves more image details in high noise levels, especially in the activity region of image.