Factorized Geometrical Autofocus for Synthetic Aperture Radar Processing英文

《Factorized Geometrical Autofocus for Synthetic Aperture Radar Processing英文》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、ThesisfortheDegreeofLicentiateofEngineeringFactorizedGeometricalAutofocusforSyntheticApertureRadarProcessingJanTorgrimssonDepartmentofEarthandSpaceSciencesChalmersUniversityofTechnologyGothenburg,Sweden2013FactorizedGeometricalAutofocusforSyntheticApertureRadarProcessingJanTo

2、rgrimssoncJanTorgrimsson,2013.TechnicalReport55L,RadarRemoteSensingGroupDepartmentofEarthandSpaceSciencesChalmersUniversityofTechnologySE{41296Gothenburg,SwedenPhone:+46(0)317721000E-mail:tojan@chalmers.sePrintedbyChalmersReproserviceChalmersUniversityofTechnologyGothenburg,S

3、weden2013FactorizedGeometricalAutofocusforSyntheticApertureRadarProcessingJanTorgrimssonDepartmentofEarthandSpaceSciencesChalmersUniversityofTechnologyAbstractSyntheticApertureRadar(SAR)imageryisaveryusefulresourceforthecivilianremotesensingcommunityandforthemilitary.Thishowe

4、verpresumesthatimagesarefocused.Thereareseveralpossiblesourcesfordefocusingeects.ForairborneSAR,motionmeasurementerrorsisthemaincause.Adefocusedimagemaybecompensatedbywayofautofocus,estimatingandcorrectingerroneousphasecomponents.Standardautofocusstrategiesareimplementedasas

5、eparatestageaftertheimageformation(stand-aloneautofocus),neglectingthegeometricalaspect.Inaddition,phaseerrorsareusuallyassumedtobespaceinvariantandconnedtoonedimension.Thecallforrelaxedrequirementsoninertialmeasurementsystemscontradictsthesecriteria,asitmayintroducespacevar

6、iantphaseerrorsintwodimensions,i.e.residualspacevariantRangeCellMigration(RCM).Thishasmotivatedthedevelopmentofanewautofocusapproach.Thetechnique,termedtheFactorizedGeometricalAutofocus(FGA)algorithm,isinprincipleaFastFactorizedBack-Projection(FFBP)realizationwithanumberofadj

7、ustable(geometry)parametersforeachfactorizationstep.Byalteringtheapertureinthetimedomain,itispossibletocorrectanarbitrary,inaccuratege-ometry.ThisinturnindicatesthattheFGAalgorithmhasthecapacitytocompensateforresidualspacevariantRCM.Inappendedpaperstheperformanceofthealgorith

8、misdemonstratedforgeometricallycon-strainedautofocusproblems.Results