图像盲恢复算法及其实现

针对图像盲恢复算法及其应用进行了研究，讨论了盲恢复算法中的非参数限定支持域恢复技术，这类算法并不假设图像或者模糊的参数模型，而是同时估计出清晰图像和点扩展函数。在介绍Kundur等提出的非负和集约束递归逆滤波算法的基础上，提出了该算法的改进，利用零相位FIR滤波器的频率特性，消除滤波器带来的相位失真，减少图像失真；增大FIR系数中心像元的加权值，增强图像高频。实验结果证明了改进算法的有效性。

The Algorithm and Implementation of Image Blind Restoration

Mainly research blind image restoration. A class of nonparametric finite support restoration techniques is illuminated in detail. The method of this class doesn't assume image or degraded image parameter module, while estimates the true image and the PSF at the same time.We introduce adaptive filtering applications on signal processing, and NAS-RIF algorithm proposed by Kundur.We suggest methods for improving NAS-RIF algorithms, which utilize frequent a characteristics of zero-phase FIR filter, remove phase distortion caused by the filter and reduce image distortion. Increasing center weighted values of FIR coefficients enhances high-frequencies of image. Experimental results indicate that these improvements are effective.