用于含噪二值图像的改进NAS-RIF图像盲复原算法

D.Kunder等提出的NAS－RIF图像盲复原算法在无噪情况下，能够得到好的结果，但算法对噪声敏感。针对此特点，Michael K.Ng等提出了NAS－RRIF算法，它主要是对逆滤波器进行正则化，以避免噪声过分放大，但 所得结果没有大的改善。二值图像的盲复原也是其应用价值的，本文将二值图像所具有的更多先验信息用于NAS－RIF算法，采用新的非线性函数，并结合后续处理,提出一种用于含噪二值图像的改进算法。实验表明，在噪声情况下，所提出的算法可以得到满意的结果。算法还考虑了背景灰度值非零情况。文章给出了实验结果，另外还考虑并解决了NAS－RIF算法固有的比例问题。

An Improved NAS-RIF Blind Image Restoration Algorithm for Noised Two-Valued Images

The NAS RIF algorithm for blind image deconvolution proposed by D. Kunder and D. Hatzinakos can obtain good image estimates when free of noise, but it cannot get a satisfying result in the presence of noise. Because of this characteristic, Michael K.Ng, etc. presented a NAS RRIF algorithm, which mainly regularizes the inverse filter to avoid the excess amplification of noise. However, NAS RRIF algorithm doesnot greatly improve the result. Blind restoration of two valued images is also important, so we combine more priori information of two valued image with NAS RIF algorithm and propose an improved NAS RIF algorithm for noised two valued images. The algorithm also considers the situation when the background is not black and overcomes the scale problem of NAS RIF algorithm.