基于多特征空間的遙感專題信息自動提取方法研究.pdf

1、東北師范大學碩士學位論文基于多特征空間的遙感專題信息自動提取方法研究姓名：楊桄申請學位級別：碩士專業(yè)：自然地理學指導教師：劉湘南2003.5.1AbstractRemotelysensedimageryiSthemostimmediateandabundantcarrierofintegratedinformationforgroundtarget，andextractingterrainandthematicinformationfr

2、omitautomaticallyisanimpendingandcomplicatedproblemforremotesensingandGISAlongwiththequickdevelopmentofremotelysensedobtainingtechnologyinouterspace，whichprovidesmoreandmoredatastudypersonnelCangetliRleeffectiveinformati

3、onsynchronouslyyetAutomaticextractionofthematicinformationfromremotelysenseddatahasbecomebottleneckinRemotelysensedinformationproductionflowExtractionmethodbasedonspectralcharacterandsingletargetfigurecharacterhasbeendev

4、elopedmanyyears，butinterpretationimpactisshortofanticipationdesiredespiteofgeeingsomeevolutionThispaperstudiedRemotelysensedinformationexpressionofobjectmanifoldrecognitioncharacters，analyzedalluderelationbetweenRemotely

5、sensedimageryandtargetmulti—charactersspace，putforwardthestudymethodofautomaticRemotelysensedthematicinformationextractionbasedonmulticharactersspaceAccordingtotheevaluationforcomthematicinformationsupervisedclassificati

6、onresult，withtheanalysisofinfluencingfactorstocomspatialdistributing，thispaperbroughtforwardautomaticextractionmodelofcornthematicinformationbasedonmulticharactersspace，builttherelevantexpenknowledgestoreroomandfinisheda

7、utomaticclassificationofTMimagerythematicinformationusingitComparingthisclassificationresultwiththemanpowervisualinterpretation，itWasfoundthattheformercouldimproveefficiencyofthematicinformationextraction，atthesametime，t

8、hismethodwassignificanttheoreticallyandalSOaninnovationinautomaticinterpretationofremotelysensedimageryThisPaDerwasdividedintofourchaptersasfcIllows：Thefirstchaptersummarizedthepresentstudyanddevelopmentdirectionaboutaut

9、omaticinterpretationofremotesensedimageryandthenexplainedthemainstudycontent，significanceaimandstudymethodinthispaper111esecondchapterexpatiatedexpressionofmanifoldRemotelysensedinformationaboutgroundtargetsbroughtforwar