一种基于凸集投影(POCS)的数字图像超分辨率重建算法  

A Super-resolution Image Reconstruction Algorithm on the Theory of Projection Onto Convex Sets(POCS)

作　　者：庞亚红[1] 毛幼菊[1] 

机构地区：[1]重庆邮电学院光通信研究所,重庆400065

出　　处：《计算机工程与应用》2005年第4期69-71,共3页Computer Engineering and Applications

基　　金：重庆市信产局产业化资助项目(编号:200113010);重庆市信产局信息化战略研究资助项目(编号:200216006)

摘　　要：该文研究了一种基于凸集投影(POCS)算法的超分辨率图像重建方法,分析了POCS方法恢复图像的理论算法,通过仿真对比了其与双线性插值方法恢复超分辨率图像的差异,仿真结果表明,该方法明显地提高了超分辨率图像的恢复质量。This paper proposes a super-resolution image reconstruction algorithm on the theory of projection onto convex sets(POCS),analyzes the theory algorithm of POCS to reconstruct image.While comparing the distinction of the results of POCS with bilinear interpolation method,authors find that POCS works are better than bilinear interpolation,such result shows that by using POCS algorithm,the quality of the super-resolution image reconstruction is significantly improved.

关 键 词：超分辨率 图像恢复 凸集投影 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]