外文翻译.docx

外文翻译.docx

《外文翻译.docx》由会员分享，可在线阅读，更多相关《外文翻译.docx（10页珍藏版）》请在冰豆网上搜索。

外文翻译

WavelettransforminimageprocessinginsimulationandApplication

1,tasksignificance

Inthetraditionalanalysisofsignalinfrequencydomain,iscompletelyunfolded,doesnotcontainanytimefrequencyinformation,whichforsomeapplicationsitisappropriate,becausethefrequencyofthesignaltoitsinformationisveryimportant.Butitsdiscardedtimeinformationmaybepossibleforsomeapplicationsalsoisveryimportant,sotheanalysisofthepromotion,putforwardalotoftimedomainandfrequencydomaininformationsignalanalysismethods,suchasshortFouriertransform,Gabortransform,time-frequencyanalysis,wavelettransform.WaveletanalysisovercomestheSTFTinasingleresolutiononthedefect,hasthecharacteristicsofmulti-resolutionanalysis,whichhasbeenwidelyappliedinimageprocessing.

Thetraditionalsignaltheory,isbuiltonthebasisoftheanalysisofFourier,Fouriertransformisakindofglobalchange,ithassomelimitations.Inpracticalapplication,thepeoplestarttoFouriertransformareimproved,thusresultinginwaveletanalysis.Waveletanalysisisanewbranchofmathematics,itisauniversalfunction,Fourieranalysis,harmonicanalysis,numericalanalysisofthemostperfectcrystalline;inthefieldsofapplication,especiallyinsignalprocessing,imageprocessing,speechprocessingandnonlinearsciencedomain,itisconsideredtobetheFourieranalysisafteranothereffectivewhenfrequencyanalysismethod.WavelettransformandFouriertransform,isatimeandfrequencydomainofthelocaltransformwhichcaneffectivelyextractedfromthesignalinformation,throughdilationandshiftoperationfunctiontofunctionorsignalmultiscaleanalysis（MultiscaleAnalysis）,tosolvetheFouriertransformcannotsolvemanydifficultproblems

Wavelettransformisarapiddevelopmentandmorepopularsignalanalysismethod,theimageprocessingisaveryimportantapplication,includingimagecompression,imagedenoising,imagefusion,imagedecomposition,imageenhancement.Waveletanalysisistheanalysismethodofthinkinginthedevelopmentandcontinuation.Inadditiontocontinuouswavelet,discretewavelettransform（CWT）（DWT）,andthewaveletpacket（WaveletPacket）andmultidimensionalwavelet

Waveletanalysisinimageprocessingapplicationsareveryimportant,includingimagecompression,imagedenoising,imagefusion,imagedecomposition,imageenhancement.Wavelettransformisanewtransformanalysismethod,ithasinheritedanddevelopedtheSTFTlocalizationofthought,andalsoovercomesthewindowsizedoesnotvarywithfrequencyandothershortcomings,toprovideafrequencychangingwithtimefrequencywindow,isatime-frequencysignalanalysisandprocessingtheidealtool.Itismainlycharacterizedbytransformcanhighlightsomeaspectsofcharacteristics,therefore,thewavelettransforminmanyareashavebeensuccessfullyapplied,especiallywavelettransformdiscretedigitalalgorithmhasbeenwidelyusedinmanyoftheproblemsofthetransformationresearch.Sincethen,thewavelettransformismoreandmoretheintroductionofpeople'sattention,itsapplicationfieldsmoreandmorewidely.

2,problemoverview

（a）theapplicationofwaveletanalysisanddevelopment

Theapplicationofwaveletanalysisandwaveletanalysistheorytoworkcloselytogether.Now,ithasbeenintheinformationtechnologyindustryhasmadetheachievementattractpeople'sattention.Electronicinformationtechnologyisthesixnewandhightechnologyanimportantfield,whichisanimportantaspectofimageandsignalprocessing.Nowadays,signalprocessinghasbecomeanimportantpartoftheworkofcontemporaryscienceandtechnology,thepurposeofsignalprocessingis:

accurateanalysis,diagnosis,codingandquantization,fasttransmissionorstorage,accuratelyreconstruct（orreturn）.Fromamathematicalperspective,signalandimageprocessingcanbeunifiedasaletter

Coursenumberprocessing（imagecanbeviewedasatwo-dimensionalsignal）,thewaveletanalysisofthemanyanalysisformanyapplications,canbeattributedtothesignalprocessingproblems.Now,foritspropertieswithtimestablesignal（stationaryrandomprocess）,anidealtoolinprocessingisstillaFourieranalysis.Butinthepracticalapplicationofthevastmajorityofsignalisunstable（nonstationaryrandomprocess）,andisespeciallysuitablefortheunstablesignalwaveletanalysistoolis.

Infactthewaveletanalysisappliedfieldisveryextensive,itincludesmanydisciplines:

mathematics;signalanalysis,imageprocessing;quantummechanics,theoreticalphysics;militaryelectronicwarfareandweaponscomputerintelligent;classificationandrecognition;musicandlanguageartificialsynthesis;medicalimaginganddiagnosis;seismicdataprocessing;mechanicalthefaultdiagnosisandsoon;forexample,inmathematics,ithasbeenusedinnumericalanalysis,structure,fastnumericalmethodofcurveandsurfacestructure,solvingdifferentialequations,controltheory.Insignalanalysis,noisefiltering,compression,transmissionandsoon.Intheimageprocessingoftheimagecompression,classification,identificationanddiagnosis,suchasthedecontamination.Inmedicalimaging,thereductionofBultrasound,CTnuclearmagneticresonanceimagingtime,improvetheresolution

（1）applicationofwaveletanalysisinsignalandimagecompressionwaveletanalysisisanimportantapplicationofthe.Itischaracterizedbyhighcompressionratio,compressionspeed,thecompressedsignalcanbemaintainedandimagefeatureinvariant,andthetransferofantiinterference.Thecompressionmethodbasedonwaveletanalysis,comparativesuccessofwaveletpacketbestbasemethod,wavelettexturemodelmethod,wavelettransformZerotreecompression,wavelettransformvectorcompression.

（2）thewaveletinthesignalanalysisarewidelyused.Itcanbeusedforboundaryprocessingandfiltering,time-frequencyanalysis,signal-noiseseparationandextractionofweaksignal,fractalindex,signalrecognitionanddiagnosisaswellasthemulti-scaleedgedetection.

Inconclusion,becausewavelethaslowentropy,multi-resolution,decorrelation,selectedmediumcharacteristicssuchasflexibility,thetheoryofwaveletindenoisingfieldsbymanyscholars,andobtainedgoodresults.Buthowtotakecertaintechnologytoeliminateimagenoisewhilepreservingimagedetailisanimportanttopicintheimagepretreatment.Atpresent,basedonwaveletanalysisinimagedenoisingimagedenoisingtechnologyhasbecomeanimportantmethod.

（b）intheimageprocessingfield,wavelettransformhasthefollowingadvantages:

（1）waveletdecompositioncancoverthewholefrequencydomain（providesamathematicallycompletedescription）

（2）wavelettransformbyselectingappropriatefilter,cangreatlyreduceorremovethecorrelationbetweendifferentfeatureextraction

（3）wavelettransformhasa"zoom"characteristics,inthelowfrequencybandcanbeusedwithhighfrequencyresolutionandlowresolution（widthanalysiswindow）,inthehighfrequencyband,theavailablelowfrequencyresolutionandhightemporalresolution（narrowanalysiswindow）

（4）wavelettransformonafastalgorithm（Mallatalgorithm）Waveletanalysishasbecomeoneofthefastestandmostattractsb.'sattentionononeofthesubjects,orappliedtothefieldofinformationinvolvedinalmostalldisciplines.

（c）demonstrationprogram

Thispaperbasedonthewavelettransformimagedenoisingmethodscarriedoutin-depthresearchandanalysis,thepaperintroducesseveralclassicwavelettransformdenoisingmethod.Thewavelettransformmodulusmaximumdenoisingmethod,describedindetailthedenoisingprincipleandalgorithm,analyzesthedenoisingprocessparameterselectionproblem,andgivessomebasis;detailedcorrelationofwaveletcoefficientdenoisingmethodprincipleandalgorithm;wavelettransformthresholddenoisingmethodprincipleandseveralkeytheproblemisdiscussedindetail.Thelastofthesemethodsareanalyzedandcompared,anddiscussestheirrespectiveadvantagesanddisadvantagesandapplicableconditions,andgivesthesimulationresults.

Inmanyimagedenoisingbasedonwavelettransformmethod,isthelargestusewaveletshrinkagedenoisingmethod.Thetraditionalhardthresholdfunctionandsoftthresholdfunctionde-noisingmethodinpracticehasbeenwidelyused,andachievedgoodresults.ButthehardthresholdfunctionisdiscontinuousresultingreconstructedsignalpronetofalsephenomenonofGibbs;andthesoftthresholdfunctionalthoughtheoverallgoodcontinuity,buttheestimatedvalueandtheactualvalueofaggregateinthepresenceofconstantdeviation,withcertainlimitations.Inviewofthis,thispaperputsforwardamethodbasedonwaveletmultiresolutionanalysisandminimummeansquareerrorcriterionforadaptivethresholddenoisingalgorithm.Themethoduseswaveletthresholdde-noisingprinciple,basedonminimummeansquareerroralgorithmofLMSandSteinunbiasedestimatesofthepremise,thederivationofacontinuousderivativewithmultiplethresholdfunction,theuseofthethresholdforiterativeoperation,gettheoptimalthreshold,resultinginbetterimagedenoisingeffect.Finally,thesimulationresultscanbeseen,thedenoisingeffectisremarkable,andhardthreshold,softthresholdmethodiscompared,theSNRimprovementmore,atthesametimedenoisingcanpreserveimagedetails,isaneffectivemethodforimagedenoising.

Waveletbasisfunctionfromthefollowing3aspectstoconsider.

（1）complexandrealwaveletselectionComplexwaveletanalysiscannotonlyobtaintheamplitudeinformation,canalsobeobtainedfromthephaseinformation,sothecomplexwavele