太赫兹 反卷积 matlab,一种基于小波的光谱反卷积方法(英文)

Inspectrummeasurement,theresolutionofareceivedsignalisgreatlyaffectedbyslitwidthofthemonochmmator[“.Ingeneral,increasingtheslitwidthwillleadtoadecreaseofresolution.Therefore。anarrowslitisalwayspreferred.Ontheotherhand,whenmeasuringalowintensityspectrum,awideslitmustbeusednotonlytoincreasethereceivedareas,butalsotogetadegradedspectrum.Theinfluenceoftheslitwidthcanbedeseribedasaslitfunction,andthereceivedspectrumisaconvolutionoftheidealspectrumandtheslitfunction.ThUS.deconvolutioniSneededtorestorethehighresolutionspectrum,especiallyinlowintensityspectrummeasurement.ThesignaldeconvolutionproblemisaninverseproblemmathematicallymodeledbyaFredholmintegralequationofthefirstkind.,11lisequationiSusuallyanillposedproblemwhenitisconsideredinaHilbertspaceframeworkcontrolarbitrary,requiringregularizationtechniquestoerroramplificationsandtogetadequatesolutions[213].Inthispaper,wepresent,underapmjectionontoconvexsets(POCS)framework,awavelet-basedmethodforobtainingaregularizedsolutiontodiscreteFredholmintegralequationsofthefirstkindcorruptedbyadditivenoise.OurmethodincludesthemuhiscaleedgesreconstructionalgorithmwhichWasrecentlyappliedtospikydeconvolution‘制.Inourcase,wedealwithamoregeneralmodel,andourmethodhasawiderapplication.Theconvolutionprocesshasthefollowinggenericform:h*,+n=f(1)wherehiStheslitfunction;nisnoise;,iStheidealspectrum;|isthereceivedspectrum。Afterdiscretization,Eq.(1)becomesHr+万=f(2)whereHisamatrixcorrespondingtotheactionoffilterh.POCS【。]widelyusedintherecoveryproblemisanite