欢迎来到天天文库
浏览记录
ID:9373352
大小:3.20 MB
页数:8页
时间:2018-04-29
《多尺度细节增强的遥感影像超分辨率重建_朱红》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第45卷第9期测绘学报Vol.45,No.92016年9月ActaGeodaeticaetCartographicaSinicaSeptember,2016引文格式:朱红,宋伟东,谭海,等.多尺度细节增强的遥感影像超分辨率重建[J].测绘学报,2016,45(9):1081-1088.DOI:10.11947/j.AGCS.2016.20150451.ZHUHong,SONGWeidong,TANHai,etal.RemoteSensingImagesSuperResolutionReconstructionBasedonMulti-scaleDetailEnhancement[J]
2、.ActaGeodaeticaetCartographicaSinica,2016,45(9):1081-1088.DOI:10.11947/j.AGCS.2016.20150451.多尺度细节增强的遥感影像超分辨率重建1,2121朱红,宋伟东,谭海,王竞雪1.辽宁工程技术大学测绘与地理科学学院,辽宁阜新123000;2.国家测绘地理信息局卫星测绘应用中心,北京100048RemoteSensingImagesSuperResolutionReconstructionBasedonMulti-scaleDetailEnhancement1,2121ZHUHong,SONGWeidon
3、g,TANHai,WANGJingxue1.SchoolofGeomatics,LiaoningTechnicalUniversity,Fuxin123000,China;2.SatelliteSurveyingandMappingApplicationCenterofNASG,Beijing100048,ChinaAbstract:Theexistingmethodsarehardtohighlightthedetailsaftersuperresolutionreconstruction,soitisproposedasuper-resolutionmodelframetoenh
4、ancethemulti-scaledetails.Firstly,thesequenceimagesaremulti-scaledeposedtokeeptheedgestructureandthedeposedmulti-scaleimageinformationaredifferenced.Then,thesmoothinginformationanddetailinformationareinterpolated,andatexturedetailenhancementfunctionisbuilttoimprovethescopeofsmalldetails.Finally
5、,thecoarse-scaleimageinformationandsmall-medium-scaleinformationareconfusedtogetthepremiersuper-resolutionreconstructionresult,andalocaloptimizingmodelisbuilttofurtherpromotethepremierimagequality.Theexperimentsonthesameperiodanddifferentperiodremotesensingimagesshowthattheobjectiveevaluationin
6、dexarebothlargelyimprovedcomparingwiththeinterpolationmethod,traditionaltotalvariation(TV)method,andmaximumaposterior(MAP)method.Thedetailsofthereconstructionimageareimproveddistinctly.Thereconstructionimageproducedusingtheproposedmethodprovidesmorehighfrequencydetails,andthemethodprovestoberob
7、ustanduniversalfordifferentkindsofsatelliteremotesensingimages.Keywords:superresolutionreconstruction;detailenhancement;multi-scaledecomposition;imagefusionFoundationsupport:ThePublicWelfareIndustryScientificResearchBasedonSurveyi
此文档下载收益归作者所有