基于MTF的遙感圖像恢復(fù)技術(shù)研究.pdf

1、南京理工大學(xué)碩士學(xué)位論文基于MTF的遙感圖像恢復(fù)技術(shù)研究姓名：邱曉君申請學(xué)位級別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：夏德深20060601■●●碩十論文摹于MTF的遙巷圖像恢復(fù)技術(shù)研究ABSTRACTTheremotesensingimagesarecapturedbythesatelliteonorbitInsomecases，wemaygettheblurredremotesensingimagesTherearemanycause

2、s，suchastheatmosphere，theremotesensors，andthesignaltransmissionSoit’SimportanttoimprovethequalityofremotesensingimagesTheremanydifferentwaysofimagerestorationHoweverlotsofthesewaysarebasedonthePSForneedotherconditionsofi

3、magingsystemSometimes，wecan’tgetenoughconditions，anditisdifficulttocalculateInthesecases，itishardtestoretheimagesbytheusualmethodsSo，inthispaperwewilldiscussadifferentmethodofimagerestorationwhichdoesnotneedtheknowledgeo

4、fthePSForotherconditionsInthispaperwedidsomeresearchasfollows：1FirstlyweremovedthenoiseoftheimagesAimatdifferentnoises，weuseddifferentmethodsForthestripnoise，thefrequencymethodbasedonFourierTransformWasusedandlocalselfad

5、aptationMedianFilterWasadoptedfortheisolatednoiseThesetwomethodscouldremovethenoiseeffectivelyatthesametime，preventtheimagesfromblurring2ThenWerestoredtheimagesTheblurredimagescouldbeimprovedbasedonthedegradationcharacte

6、risticofMTFinfrequencyareaFirst，thedegradationmodelWasestabSshedbasedonMTFSecondtransformedtheimagesfromspatialareatofrequencyareabyFourierTransformAndthen，restoredtheimagesbasedonthedegradationmodelinfrequencyareaAtlast

7、，transformedtheimagestospatialareaBythiswaywecouldimprovetheimagesinadegree3Also，thispaperdiscussedtherelationsbetweenMTFandPSI：andestablishedPSFmodelbasedonMTFDifferenttypesofPSFincluding1Dand2DPSFwerealldiscussedinthee

8、xperimentinordertoacquirethesuitablePSFmodelAndweadoptedthisPSFmodelinsomeusualimagerestorationalgorithmsTheresultsprovedthatthismethodCangetbetterimagesThemethodsintiffspaperaresuitablenotonlyforremotesensingimages，buta