预览加载中，请您耐心等待几秒...

非负矩阵分解在遥感图像变化检测中的应用研究的开题报告.docx

预览
1/2
2/2

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

非负矩阵分解在遥感图像变化检测中的应用研究的开题报告 Title:ApplicationofNon-negativeMatrixFactorizationinChangeDetectionofRemoteSensingImages Background: RemotesensingisanimportanttoolformonitoringchangesintheEarth'ssurface.Changedetectionofremotesensingimagesreferstotheprocessofdetectingchangesthathaveoccurredinthesurfacefeaturesofagivenareainsuccessiveimagestakenatdifferenttimes.Itiswidelyusedinfieldssuchasdisasterassessment,urbanplanning,environmentalmonitoring,andmilitaryapplications.Traditionalchangedetectionmethodsoftenrelyonmanualinterpretationorsupervisedalgorithms,whicharetime-consumingandsubjective.Moreover,thesemethodsrequirehigh-qualityimageswithoutnoise,andtheirperformancedeteriorateswhendealingwithcomplexscenes. Non-negativematrixfactorization(NMF)isapopularunsupervisedlearningtechniqueusedfordataanalysis,particularlyinimageprocessing.Itminesusefulinformationbydecomposinganon-negativematrixintotwonon-negativefactorssuchthattheproductofthetwofactorsapproximatestheoriginalmatrix.Thisfactorizationisunique,andtheresultingfactorscapturetheinherentfeaturesoftheinputdata.Duetoitsabilitytoextracthiddenfeatures,NMFhasbeensuccessfullyappliedtovariousimageprocessingtasks,suchasimageclustering,imagesegmentation,andimageclassification. ResearchAim: TheaimofthisresearchistoinvestigatetheeffectivenessofNMFinchangedetectionofremotesensingimages.Specifically,wewillexplorethefollowingresearchquestions: 1.CanNMFaccuratelydetectchangesinremotesensingimages? 2.HowdoestheselectionofparametersaffecttheperformanceofNMFinchangedetection? 3.HowdoesNMFcomparewithtraditionalchangedetectionmethodsintermsofaccuracyandcomputationalefficiency? Methodology: Wewillconductexperimentsonpubliclyavailableremotesensingdatasets,suchasLandsatandSentinel-2,toevaluatetheperformanceofNMFinchangedetection.WewillcomparetheresultsofNMFwithtraditionalchangedetectionmethods,suchasthepost-classificationmethodandimagedifferencemethod.Wewillconsidervariousparameters,suchasthenumberofcomponents,thechoiceofinitializationmethod,and