基于邻域信息的图像自适应滤波

Adaptive Filtering Based on Neighborhood Information

该文分析了常规图像滤波算法的局限性,并在此基础上提出了一种基于邻域信息的自适应滤波新算法。该算法利用噪声的灰度不一致性,通过邻域信息差分值的差异来决定象素的类别,然后针对不同类别的象素点采取不同的滤波算法。实验结果表明,该算法能够显著提高图像的信噪比。

This paper analyzed the localization of general filtering algorithm, and presented a new adaptive filtering algorithm based on neighborhood information. This algorithm made use of the variance of noise's gray-value, divided pixels to different kinds by the difference of neighborhood information, and then adopted different dispose to different kinds of pixels. The experiment result shows that this algorithm can improve the signal-to-noise ratio(SNR) greatly.