一种基于权值矩阵的序列图像超分辨率盲复原算法  

A Weight Matrix Based Blind Super Resolution Restoration Algorithm

作　　者：王素玉[1] 沈兰荪[1] 卓力[1] 李晓光[1] 

机构地区：[1]北京工业大学信号与信息处理研究室,北京100124

出　　处：《电子学报》2009年第6期1198-1202,1209,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.6047203660772069);北京市自然科学基金(No.4052007);北京市科技新星计划(No.2005B08)

摘　　要：建立准确的观测模型是提高现有超分辨率复原算法性能的关键.本文提出一种基于权值矩阵的超分辨率盲复原算法:定义了一种新的基于权值矩阵和运动补偿矩阵的观测模型,在最大后验概率框架下,采用交替最小化方法对权值矩阵和高分辨率图像进行联合优化求解.静止和动态图像序列的测试结果表明,该方法能够实现对低分辨率图像降质过程的准确描述,其复原性能明显优于传统基于理想观测模型的算法,部分结果甚至超过了观测模型已知的算法.The observation model plays a key role in performance improvement of the super resolution algorithms. The author proposed a weight-matrix based blind super resolution algorithm： a new observation model based on a motion compensate matrix and a weight-matrix is defined first, then the high resolution image and the weight matrix were joint estimated by alter minimization method under the farmework of Maximum A Prior（MAP）. Evaluated by both still and active image sequences, the algorithm can descript the degrading process of the observed low resolution images more accurately. It shows obvious performance improvement compared with the traditional super resolution algorithm. For several cases, it has even exceeded the results when the observation model is known.

关 键 词：超分辨率盲复原 观测模型 最大后验概率 权值矩阵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]