|本期目录/Table of Contents|

[1]鲁林,李翠华*,张珍,等.一种MRFMAP框架下的图像超分辨率重建方法[J].厦门大学学报(自然科学版),2012,51(4):696.
 LU Lin,LI Cui hua*,ZHANG Zhen,et al.A Method of Images Superresolution Reconstruction Based onMRFMAP Frame[J].Journal of Xiamen University(Natural Science),2012,51(4):696.
点击复制

一种MRFMAP框架下的图像超分辨率重建方法(PDF)
分享到:

《厦门大学学报(自然科学版)》[ISSN:0438-0479/CN:35-1070/N]

卷:
51卷
期数:
2012年第4期
页码:
696
栏目:
研究论文
出版日期:
2012-07-15

文章信息/Info

Title:
A Method of Images Superresolution Reconstruction Based on MRFMAP Frame
作者:
鲁林李翠华*张珍余礼钹张东晓施华
厦门大学信息科学与技术学院,福建 厦门 361005
Author(s):
LU LinLI Cuihua*ZHANG ZhenYU LiboZHANG DongxiaoSHI Hua
School of Information Science and Technology,Xiamen University,Xiamen 361005,China
关键词:
超分辨率重建马尔科夫随机场最大后验概率图切算法
Keywords:
superresolution reconstructionmarkov random field(MRF)maximum a posteriori(MAP)graphcut
分类号:
TP 391.4
文献标志码:
-
摘要:
基于多帧观察样本的超分辨率图像重建是超分辨率图像重建研究中的重要方向.在马尔科夫随机场最大后验概率(MRFMAP)框架下研究了多帧图像的超分辨率重建问题.根据给定的空间图像退化模型建立了超分辨率重建的二阶能量函数,并利用αexpansion图切算法对能量函数进行求解.考虑到αexpansion算法的规范性要求,将能量函数进行近似.针对二阶能量函数的图切算法,讨论了st图的构造,给出一种节点的分配方法以及tlink和nlink的赋值方式,以提高图切算法的计算效果.通过对两种类型的图像进行超分辨
Abstract:
Super resoltion based on multiobservation play a impostant role in the studing of subarresoluting.This paper focus on superresolution on images based on the frame of MRFMAP.We construct the two clique energy function according to the images observation model.After a brief introduction to graphcut,we use αexpansion algorithm to calculate the optima of energy function.Considering the regularity that αexpansion algorithm have on energy function,we approximate the energy function.In order to improve the quality of the reconstruction result,we design the construction of st graph for graphcut algorithm and the assignment of the weight of the tlink and nlink.Via comparative experiment using two different type of pictures for superresolution reconstruction,the method in this paper do better on superresolution and denoising.

参考文献/References:

[1]Patti A J,Altunbasak Y.Artifact reduction for set theoretic super resolution mage reconstruction with edge adaptive constraints and higherorder interpolants[J].IEEE Transactions on Image Processing,2001,10(1):179186. [2]Yang J,Wright J,Huang T S,et al.Image superresolution via sparse representation[J].IEEE Transactions on Image Processing,2001,10(1):113. [3]Borman S,Stevenson R.Linear models for multiframe superresolution restoration under nonaffine registration and spatially varying PSF[J].SPIE,2004,5299:234245. [4]Irani M,Peleg S.Motion analysis for image enhancement: resolution,occlusion and transparency[J].Journal Visual Communication and Image Representation,1993,4(4):324335. [5]Spitzer F.Markov random fields and Gibbs ensembles[J].The American Mathematical Monthly,1971,78(2):142154. [6]Pelillo M.The dynamics of nonlinear relaxation labeling processes[J].Journal of Mathematical Imaging and Vision,1997,7:309323. [7]Greig D,Porteous B,Seheult A.Exact maximum a posteriori estimation for binary images[J].Journal of the Royal Statistical Society:Series B,1989(2):271279. [8]Ishikawa H,Geiger D.Segmentation by grouping junctions[C]//IEEE Conference on Computer Vision and Pattern Recognition.New York: IEEE Press,1998:125131. [9]Boykov Y,Veksler O,Zabih R.Markov random felds with effcient approximations[C]//IEEE Conference on Computer Vision and Pattern Recognition.CA:IEEE,1998:648655. [10]Boykov Y,Veksler O,Zabih R.Fast approximate energy minimization via graph cuts[J].IEEE Transactionson Pattern Analysis and Machine Intelligence,2001,23(11):12221239. [11]Boykov Y,Kolmogorov V.An experimental comparison of mincut/maxflow algorithms for energy minimizationin vision[J].IEEE Transactions Pattern Analysis and Machine Intelligence,2004,26(9):11241137. [12]Kolmogorov V,Zabih R.What energy functions can be minimized via graph cuts[J].IEEE Transactionson Pattern Analysis and Machine Intelligence,2004,26(2):147159. [13]Mudenagudi U,Banerjee S,Kalra P K.Spacetime superresolution using graphcut optimization[J].IEEE Transactionson Pattern Analysis and Machine Intelligence,2011,33(5):9951008. [14]Li S Z.Markov random field modeling in image analysis[M].3rd ed.Berlin:Springer,2009:2128. [15]Baker S,Kanade T.Limits on superresolution and how to break them [J].IEEE Computer Vision and Pattern Recognition,2002,24(9):11671183.

备注/Memo

备注/Memo:
收稿日期:20111231基金项目:国防基础科研计划项目;国防科技重点实验室基金;福建省自然科学基金项目(2011J01365);高等学校博士学科点专项科研基金项目(20110121110020)*通信作者:chli@xmu.edu.cn
更新日期/Last Update: 2012-07-15