2009_超分辨率算法研究综述_浦剑.pdfVIP

39 1 ( ) 2009 2 Vol . 39 No. 1 JOURNAL OF SHANDONG UNIVERSITY ( ENGINEERING SCIENCE) Feb . 2009 : 1672- 396 1( 2009) 01- 0027- 06 1 1* 2 浦剑 , 张军平 , 黄华 ( 1. , 200433; 2. , 710049) : 图像超分辨率是指利用 一幅或多幅低分辨率图像, 运用相应的算法来获得 一幅 晰的高分辨率图像. 然而, 传统的基于插值和重建的方法已很难获得进 一步的突破. 近年来出现的基于学习的方法为超分辨率的发展重新 注入了活力. 通过回顾插值重建和学习这 3 个层面的超分辨率算法, 分析了超分辨率技术的以往研究和最新进 展, 着重讨论了各算法在还原质量通用能力等方面所存在的问题, 并对未来超分辨率技术的发展作了 一些展望. : 图像处理; 超分辨率; 邻域嵌入; 图像重建 : TP75111 : A A survey of super resolution algorithms 1 1* 2 PU Jian , ZHANG Jun-ping , HUANG Hua ( 1. School of Computer Science, Fudan University, Shanghai 200433, China; 2 . School of Electronics and Information Engineering, Xi. an Jiaotong University, Xi. an 710049, China) Abstract: T he goal of super resolution is to get high resolution images from at least one low resolution image. How ever, w ith the traditional ( interpolation- based and reconstruction-based) methods it is diff icult to make further important progress . The emerging learning- based super- resolution methods breathe new lif e into the research . By reviewing these three kinds of methods, the history of super resolution w as surveyed, the limitation of existing methods w as discussed and the roadmap was analyzed for future devel- opments . Key words: image processing; super resolution; neighbor embedding; image reconstruction ( super resolution , SR) ( lo