正则化数字图象盲恢复的一种新方法  

A NEW APPROACH TO REGULARIZED BLIND RESTORATION OF DIGITAL IMAGES

作　　者：罗忠[1] 朱重光[1] 

机构地区：[1]中国科学院遥感应用研究所,北京100101

出　　处：《测试技术学报》1999年第2期67-75,共9页Journal of Test and Measurement Technology

摘　　要：这种新的数字图象盲恢复的方法,是以解决病态数学问题的正则化技术为基础的,通过构造适当的正则项,将未知清晰图象和未知点扩散函数结合到一个代价泛函中,然后采用数学优化方法对代价泛函进行最小化,以达到同时获取未知清晰图象和辨识未知点扩散函数的目的.此算法可以直接对二维图象进行操作,而不必进行通常的图象矩阵到向量的相互转换,这种方法可以用于一般图象、遥感图象和数字化照片的盲恢复.A new approach to blind restoration of digital images is investigated in this paper. On the basis of the regularization technique for dealing with ill-posed mathematical problems, the unknown clear image as well as the unknown point-spread function are incorporated in a single cost functional by constructing approp-tiate regularization terms. The cost functional then is minimized so that the unknown clear image can be obtained and the unknown point-spread function can be identified simutaneously. The new algorithm can be applied directly to images without the mutual conversions between image matrices and vectors. We apply the new method to blind restoration of ordinary images, remote sensing images and digitized photographs.

关 键 词：图象盲恢复 点扩散函数 正则化 数学问题 优化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]