机构地区: 华南理工大学计算机科学与工程学院
出 处: 《暨南大学学报(自然科学与医学版)》 2007年第5期461-465,共5页
摘 要: 超分辨率图像重构是利用多帧低分辨率图像重构出一幅具有更高分辨率图像.一般的凸集投影算法在放大倍数上升时存在两个问题:一是计算复杂性急剧上升,二是边缘振荡效应的加剧导致成像质量迅速恶化.本文针对凸集投影算法,提出了一种基于约束边界模式的算法.实验结果表明,新算法能够在有效抑制边缘振荡效应的同时,较大地提高了重构速度. Super-resolution image reconstruction produces high-resolution image from a set of shifted, blurred and decimated versions thereof. However, known POCS super-resolution algorithms get worse either in computational complexity or ringing artifacts as magnification increases. A pattern-based algorithm is proposed. Experiment results demonstrate that is more efficient and can suppress ringing artifacts quite well.