简介:Specklefilteringofsyntheticapertureradar(SAR)imageswhilepreservingthespatialsignalvariability(textureandfinestructures)stillremainsachallenge.ManyalgorithmshavebeenproposedfortheSARimagerydespeckling.However,simulatedannealing(SA)methodisoneofexcellentchoicescurrently.AcriticalprobleminthestudyonSAistoprovideappropriatecoolingschedulesthatensurefastconvergencetonear-optimalsolutions.Thispapergivesanewnecessaryandsufficientconditionforthecoolingschedulesothatthealgorithmstateconvergesinallprobabilitytothesetofgloballyminimumcoststates.Moreover,itconstructsanappropriateobjectivefunctionforSARimagedespeckling.AnexperimentalresultoftheactualSARimageprocessingisobtained.
简介:ThereisdifficultyfordistinguishingofriverandshadowinSyntheticApertureRadar(SAR)images.AmethodofriversegmentationinSARimagesbasedonwaveletenergyandgradientisproposedinthispaper.Itmainlyincludestwoalgorithms:coarsesegmentationandrefinedsegmen-tation.Firstly,Theriverregionsarecoarselysegmentedbythewaveletenergyfeature,andthenrefinedsegmentedaccuratelybythegradientthresholdwhichisgotadaptively.Theexperimentalresultsshowthevalidityofthemethod,whichprovidesagoodfoundationfortargetsdetectionabovetheriver.
简介:ThemathematicalmodelofspaceborneSARsystemsanditscomputersimulationsaredescribed.Resultsofcomputersimulationsaboutrangemigration,rangemigrationcorrection,azimuthweightingandsignalgeneration/processingaregiven.ThissoftwarecanbeusedtosimulatethedynamicprocessesinspaceborneSARsystems,todevelopnewsignalprocessingtechniquesandtoevaluatetheperformanceofthedesignedsystem.
简介:ThispaperstudiesSyntheticApertureRadar(SAR)imagescontaminatedbythecoherentspecklenoisewiththemultiresolutionanalysisofwavelettransform.ThisstudyshowsthattheinfluencesofthespeckleondifferentfrequencycomponentsoftheSARimagearedifferent,andthattheSARimageandthespecklehavedifferentmannersofsingularity.So,thispaperpresentsadenoisingmethodofwaveletanalysistoreducethespeckle.Someexperimentsapprovethatthismethodnotonlysuppressesthespeckleeffectively,butalsopreservesasmuchtargetcharacteristicsoftheoriginalimageaspossible.ItshowsthatthisdenoisingmethodofwaveletanalysisoffersaveryattractivealternativetosuppressthecoherentspecklenoiseoftheSARimage.
简介:Whentheclassicalconstantfalse-alarmrate(CFAR)combinedwithfuzzyC-means(FCM)algorithmisappliedtotargetdetectioninsyntheticapertureradar(SAR)imageswithcomplexbackground,CFARrequiresblock-by-blockestimationofcluttermodelsandFCMclusteringconvergestolocaloptimum.Toaddresstheseproblems,thispaperpro-posesanewdetectionalgorithm:knowledge-basedcombinedwithimprovedgeneticalgorithm-fuzzyC-means(GA-FCM)algorithm.Firstly,thealgorithmtakestargetregion’smaximumandaverageintensity,area,lengthoflongaxisandlong-to-shortaxisratiooftheexternalellipseasfactorswhichinfluencethetargetappearingprobability.Theknowledge-baseddetectionalgorithmcanproducepreprocessresultswithouttheneedofestimationofcluttermodelsasCFARdoes.AfterwardtheGA-FCMalgorithmisimprovedtoclusterpre-processresults.IthasadvantagesofincorporatingglobaloptimizingabilityofGAandlocaloptimizingabilityofFCM,whichwillfurthereliminatefalsealarmsandgetbetterresults.TheeffectivenessoftheproposedtechniqueisexperimentallyvalidatedwithrealSARimages.
简介:ThisletterstudiesonthedetectionoftexturefeaturesinSyntheticApertureRadar(SAR)images.ThroughanalyzingthefeaturedetectionmethodproposedbyLopes,animprovedtexturedetectionmethodisproposed,whichcannotonlydetecttheedgeandlinesbutalsoavoidstretchingedgeandsuppressinglinesoftheformeralgorithm.ExperimentalresultswithbothsimulatedandrealSARimagesverifytheadvantageandpracticabilityoftheimprovedmethod.
简介:MultistageVectorQuantization(MSVQ)canachieveverylowencodingandstoragecomplexityincomparisontounstructuredvectorquantization.However,theconventionalMSVQissuboptimalwithrespecttotheoverallperformancemeasure.Thispaperproposesanewtechnologytodesignthedecodercodebook,whichisdifferentfromtheencodercodebooktooptimisetheoverallperformance.Theperformanceimprovementisachievedwithnoeffectonencodingcomplexity,bothstorageandtimeconsuming,butamodestincreaseinstoragecomplexityofdecoder.
简介:OptimizationofantennaarraypatternusedinaspaceborneSyntheticApertureRadar(SAR)systemisconsideredinthisstudy.Arobustevolutionaryalgorithm,Non-dominatedSortingGeneticAlgorithms(theimprovedNSGA-Ⅱ),isappliedonaspaceborneSARantennapatterndesign.Thesystemconsistsoftwoobjectivefunctionswithtwoconstraints.Paretofrontsaregeneratedasaresultofmulti-objectiveoptimization.AfterbeingvalidatedbyatestproblemZDT4,thealgorithmsareusedtosynthesizespaceborneSARantennaradiationpattern.ThegoodresultswithlowAmbi-guity-to-SignalRatio(ASR)andhighdirectivityareobtainedinthepaper.
简介:SyntheticApertureRadar(SAR)imagingofoceansurfacefeaturesisstudied.ThesimulationoftheturbulentandvorticalfeaturesgeneratedbyamovingshipandSARimagingofthesewakesiscarriedout.Theturbulentwakedampingtheoceansurfacecapillarywavesmaybepartiallyresponsibleforthesuppressionofsurfacewavesneartheshiptrack.Thevortexpairgeneratingachangeinthelateralflowfieldbehindtheshipmaybepartiallyresponsibleforanenhancementofthewavesneartheedgesofthesmootharea.Thesehydrodynamicphenomenaaswellasthechangesofradarbackscattergeneratedbyturbulenceandvortexaresimulated.AnSARimagingmodelisthenusedonsuchoceansurfacefeaturestoprovideSARimages.Comparisonoftwoships'simulatedSARimagesshowsthatthewakefeaturesaredifferentforvariousshipparameters.
简介:Wavelet-fractalbasedSAR(syntheticapertureradar)imageprocessingisoneoftheadvancedtechnologiesinimageprocessing.Themainconceptofanalysisisthatafterwavelettransformation,multifractalspectrumofthesignalisdifferentfromthatofnoise.ThisdifferenceisusedtoalleviatethenoiseproducedbySARimage.ThemethodtodenoiseSARimageusingtheprocessbasedonwavelet-fractalanalysisisdiscussedindetail.Essentially,thepresentmethodfocusesonadjustingtheHlderexponentαofmultifractalspectrum.Aftersimulation,αshouldbeadjustedto1.72-1.73.Themorethevalueofαexceeds1.73,thelessdistinctivetheedgesofSARimagebecome.Accordingtotheauthorsdenoisingisoptimalatα=1.72-1.73.Inotherwords,whenα=1.72-1.73,asmoothanddenoisedSARimageisproduced.
简介:Withamultiplicativespecklemodel,thispapershowsthemulti-lookpolarimetricsyntheticapertureradar(SAR)dataobeysageneralizedK-distribution.Tovalidatethisdistributionmodel,themulti-lookintensityK-distributionisparticularlytested.Therelationshipbetweentheheterogeneitycoefficientofthesceneandtheproperstatisticalmodelisexperimentallyestablished.Inaddition,basedontheresultsofthestatisticalanalysis,anadaptiveclassificationschemeispresented,andtheimprovedclassificationshowstheimportanceofthestatisticalanalysis.
简介:Todealwiththenon-Gaussiannoiseinstandard2-DSARimages,thederampedsignalinimagingplane,andthepossiblesymmetricdistributionofcomplexnoise,thefourth-ordercumulantofcomplexprocessisintroducedintoSARtomography.WiththeestimatedARparametersofARMAmodelofnoisethroughYule-Walkerequation,thesignalseriesofheightispre-filtered.Then,throughESPRIT,thespectrumisobtainedandtheapertureinheightdirectionissynthesized.Finally,theSARtomographyimagingofsceneisachieved.Theresultsofprocessingonsignalwithnon-Gaussiannoisedemonstratetherobustnessoftheproposedmethod.Thetomographyimagingofthescenesshowsthatthehigher-orderspectrumanalysisisfeasibleintheapplication.
简介:Bistaticforward-lookingsyntheticapertureradar(SAR)hasmanyadvantagesandapplicationsowingtoitstwodimensionalimagingcapability.Therecouldbevariousimagingconfigurationsbecauseofthegeometricflexibilityofbistaticplatforms,resultinginkindsofmodelsbuiltindependentlyamongwhichtherecouldbesomesimilareventhesamemotionfeatures.Comprehensiveresearchonsuchsystemsinamorecomprehensiveandgeneralpointofviewisrequiredtoaddresstheirdifferenceandconsistency.PropertyanalysisofbistaticforwardlookingSARwitharbitrarygeometryisachievedincludingstripmapandspotlightmodesonairborneplatform,missile-borneplatform,andhybridplatformofboth.Emphasisisplacedonazimuthspacevarianceofsomekeyparameterssignificantlyaffectingthesubsequentimagingprocessing,basedonwhichthefrequencyspectraarefurtherdescribedandcomparedconsideringrespectivefeaturesofdifferentplatformsforfrequencyimagingalgorithmdeveloping.Simulationresultsconfirmtheeffectivenessandcorrectnessofouranalysis.
简介:ThepaperpresentsanalgorithmofautomatictargetdetectioninSyntheticApertureRadar(SAR)imagesbasedonMaximumAPosteriori(MAP).Thealgorithmisdividedintothreesteps.First,itemploysGaussianmixturedistributiontoapproximateandestimatemulti-modalhistogramofSARimage.Then,basedontheprincipleofMAP,whenaprioriprobabilityisbothunknownandlearnedrespectively,thesamplepixelsareclassifiedintodifferentclassesc={target,shadow,background}.Last,itcomparestheresultsoftwodifferenttargetdetections.Simulationresultspreferablyindicatethatthepresentedalgorithmisfastandrobust,withthelearnedaprioriprobability,anapproachtotargetdetectionisreliableandpromising.
简介:Thecoherenceisameasurefortheaccuracyoftheinterferometricphase,andthesyntheticapertureradar(SAR)interferometriccoherenceisaffectedbyseveralsourcesofthedecorrelationnoise.ForthecircularSAR(CSAR)imaginggeometry,thesystemresponsefunctionisintheformoftheBesselfunctionwhichbringsahighsidelobe,andthehighsidelobeofCSARwillbeanimportantfactorinfluencingtheinterferometriccoherence.Theeffectofthehighsidelobeonthecoherenceisanalyzedanddeduced.BasedontheinterferometriccharacteristicsoftheslightdifferenceintheviewinganglesandthepotentialpixeloffsetintheinterferometricSAR(InSAR)images,arelationbetweentheradarimpulseresponseandthecoherencelossfunctionisderived.Fromtherelationalmodel,thecoherencelossfunctionduetothehighsidelobeofCSARisthendeduced,andcomparedwiththatoftheconventionalSAR.ItisshownthatthehighsidelobeofCSARfocusingsignalwillseverelyaffectthebaselinedecorrelationandcoregistrationdecorrelation.SimulationresultsconfirmthetheoreticalanalysisandquantitativelyshowthebaselineandcoregistrationdecorrelationdegradationduetothehighsidelobesofCSAR.
简介:Range-Doppler(RD)methodandReverse-Range-Doppler(RRD)methodarecombinedtogethertoachieveautomaticgeocodingofSyntheticApertureRadar(SAR)imagequicklyandaccuratelyinthepaper.TheRDmethodisfirstlyusedtolocatethefourcornersoftheimage,thentheotherpixelsoftheimagecanbelocatedbyReverse-Range-Doppler(RRD)method.Resamplingisperformedatlast.Theapproachhasanadvantageoverprevioustechniquesinthatitdoesnotrequiregroundcontrolpointsandisindependentofspacecraftattitudeknowledgeorcontrol.ItcancompensatetheshiftduetotheassumedDopplerfrequencyinSARimagepreprocessing.RRDsimplifiestheprocessofRD,thereforespeedsupthecomputation.TheexperimentalresultsshowthataSARimagecanbeautomatedgeocodedin30susingthesingleCPU(3GHz)with1Gmemoryandanaccuracyof10misattainablewiththismethod.
简介:ConventionalSyntheticApertureRadar(SAR)systemscannotobtainhigh-resolutionandwide-swathilluminationareaduetothewell-knownminimumantennaareaconstraint.SinglePhaseCenterMultiBeam(SPCMB)techniquecanovercomethislimitationbyaddingspatialsamplingthroughmultiplereceiversinazimuthdirection.Unfortunately,thisapproachwillleadtoanincreaseofazimuthambiguities(interbeamambiguities),becauseeachreceivebeam’smainlobeoverlapswiththeotherones’sidelobes.ThispaperprovesthatthefrontpartofSPCMBSARsystemscanbeconsideredtobeahybridfilterbank.Therefore,theazimuthsignalcanbereconstructedandtheinterbeamam-biguitiescanbeeffectivelysuppressedbyawell-designedhybridfilterbank.