基于NAS-RIF的图像盲复原算法的改进  

Improvement on the technology of image blind restoration based on NAS-RIF

作　　者：曲振峰[1] 

机构地区：[1]洛阳理工学院电气工程与自动化系,河南洛阳471023

出　　处：《郑州轻工业学院学报（自然科学版）》2010年第3期98-101,共4页Journal of Zhengzhou University of Light Industry:Natural Science

摘　　要：针对基于非负和有限支持域的递归逆滤波器的盲图像复原算法(即NAS-RIF算法)对噪声敏感问题,提出通过高阶累积量(HOS)去除噪声的方法,并在算法的迭代过程中采用图像分割技术进一步精确图像的目标支持域,提高了该算法的应用范围.仿真实验结果表明,改进后的方法取得了较好的复原效果.For support based on non-negativity and support constraints recursive inverse filtering algorithm（NAS-RIF） on the noise-sensitive issues,a higher-order statistics HOS method of removing noise was put up,and the iteration process image segmentation techniques was used to further support the exact image of the target domain and improve the application of the algorithm.Simulation results showed that the improved method obtains a better recovery results.

关 键 词：数字图像处理 图像盲复原 高阶统计量 NAS-RIF 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]