SAR成像目標(biāo)識(shí)別子區(qū)選取與匹配方法研究.pdf

1、華中科技大學(xué)碩士學(xué)位論文SAR成像目標(biāo)識(shí)別子區(qū)選取與匹配方法研究姓名：黃偉麟申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：模式識(shí)別與智能系統(tǒng)指導(dǎo)教師：楊衛(wèi)東20110120II華中科技大學(xué)碩士學(xué)位論文AbstractScenematchingareaionisatechniquethatssomereferenceareasintheflightpathaccdingtocertainrequirementscriteria.Theseareaswhichc

2、ontainobviousfeaturerichinfmationareofteneasytomatchusedtoguidetheaircrafttoscheduledtargetposition.Howtobuildacriteriathatholdsthecapabilityofflexibilityjammingresistanceisthekeypointftheionofsuitablematchingarea.Nowada

3、ysmanyaircraftsnavigateautomaticallybasedonopticalsensswhichareeasilyinfluencedbybadweatherlightingcondition.Therefetheguidancesystemwkswithalowefficiency.ToovercomethisshtcomingSyntheticApertureRadar(SAR)isusedmemeoften

4、.Asanactiveimagingsens，SARdoesnotonlyprovideimageswhichresolutionareashighasopticalones’butalsowkeffectivelywithoutbeinginfluencedbybadweathercondition.BecauseoftheseadvantagesSARisaveryimptantimagingsensfunmannedaircraf

5、tnavigationguidance.TheceproblemofSARsubimageionfmissionplanningofunmannedaircraftnavigatedbyscenematchingwasstudieddetailedlyinthisdissertation.anewmethodbasedonmachinelearningispresented.Themaincontentsofthismethodincl

6、ud(1)principleofscenematchingsystem(2)SARimagefeatureanalysis(3)theyofSupptingVectMachine(4)ingsubimageftargetrecognitionbasedonSVM.BesidesthispaperstudiedsomemathchingmethodssuitablefSARimageadoptedthemultisubimagematch