作　　者： ; ;

机构地区： 华南理工大学计算机科学与工程学院

出　　处： 《计算机工程与应用》 2007年第16期4-6,25,共4页

摘　　要： 超分辨率图像重构是将多帧低分辨率图像重构成一幅高分辨率图像的过程。一般的超分辨率图像重构算法在小放大倍数时效果比较好,但当放大倍数上升时,存在两个问题:一个是计算复杂性的急剧上升,另一个是边缘振荡效应的加剧导致成像质量迅速恶化。提出了一个能同时解决这两个问题的基于POCS的混合递归重构算法,实验结果表明,该算法能够在提高重构图像质量的同时,较大地提高重构速度。 Super-resolution image reconstruction produces a high-resolution image from a set of shifted, blurred, and decimated versions thereof. Previously published techniques usually perform well on small magnifications, however, as magnification increas- es, they suffer two severe problems： one is the exponentially increased computational complexity, the other is the degradation of reconstruction quality caused by severe ringing artifacts.APOCS-based hybrid recursive reconstruction algorithm is proposed to tackle these two problems. Experimental results demonstrate that the new approach is more efficient and can provide much better reconstruction quality.

关 键 词： 超分辨率 边缘振荡效应 计算复杂性

领　　域： [电子电信] [电子电信]