计算机应用与软件2013，Vol.30Issue(6)：4-7,18,5.DOI:10.3969/j.issn.1000-386x.2013.06.002
基于稀疏表示的差分图像估计方法研究
ON DIFFERENCE IMAGE ESTIMATION BASED ON SPARSE REPRESENTATION
摘要
Abstract
Difference images can show the changes in object scene together with the time.In this paper,we use an image feature-based method to get the difference images from a set of compressive measurement data of the object scene.The method mainly includes two steps.First it designs the optimal sampling matrix for getting compressive measurement data.Secondly,it uses closed-loop iterative technique for estimating the difference images,including the 12 and l1-based techniques.Employing the technique of 12-based paradigm can directly estimate the difference image from compressive measurements without the initial reconstruction of object scene.This method mainly exploits the spatial and temporal correlations of the object scene at two consecutive time instants.The 11-based paradigm technique mainly adopts a kind of modified total variation method and the method of basis pursuit denoising.Simulation demonstrates the efficacy of these techniques.关键词
压缩感知/差分图像/全变差/基追踪Key words
Compressive sensing / Difference image / Total variation / Basis pursuit分类
信息技术与安全科学引用本文复制引用
陈平生,龙丹..基于稀疏表示的差分图像估计方法研究[J].计算机应用与软件,2013,30(6):4-7,18,5.基金项目
国家自然科学基金项目(30900358/C100701) （30900358/C100701）
浙江省教育厅科研项目(Y2011122724) （Y2011122724）
浙江省新世纪高等教育教学改革研究项目(yb09138) （yb09138）
