搜索
421Vol．42No．120121LASER＆INFRAREDJanuary20121001-5078201201-0105-06··121121．1300332．100039、、。。Matlablean。TP391．41ADOI10．3969/j．issn．1001-5078．2012．01．023DeepstudyonwaveletthresholdmethodforimagenoiseremovingCHENXiao-xi12WANGYan-jie1LIULian121．ChangchunInstituteofOpticsFineMechanicsandPhysicsChineseAcademyofSciencesChangchun130031China2．GraduateUniversityofChineseAcademyofSciencesBeijing100039ChinaAbstractInviewofthedrawbacksoftraditionalwaveletthresholdmethodforimagenoisereductionthatitcannotre-movethenoisethoroughlyandmayremovetheusefulinformationthisarticleimprovesthethresholdselectionandthresholdfunctionwhicharethevitalfactorsofthismethod．Thisarticlewillimprovetheunifiedthresholdtomakethethresholdvarywiththechangesofthedecompositionscale．Atthesametimethedeviationofwaveletcoefficientando-riginalcoefficientwilldecreasethereforethemethodwillretaintheadvantagesofthetraditionalsoftthresholdandhardthreshold．Theproposedmethodwithnewthresholdselectionandthresholdfunctionwillmakeitmoreflexibletodealwiththewaveletcoefficientintheprocessofimagenoisereductionwiththehelpofwaveletdecomposing．Experi-mentresultswithMatlabrevealthatthistechnologycanlargelyreducethenoiseoftheCameraimageaddedwithGaussiannoiseandtheimagecontrastSentropyandSNRsignaltonoiseratioalsogetimproved．Theeffective-nessandprecisionofthenewwaveletthresholdmethodforimagenoisereductionpresentedinthisarticlearefinelyprovedbytheresultsoftheexperiments．Keywordsimageenhancementimagenoisereductionwavelettransformationthresholdexecute1。No．60902067。1987－。E-mail1075163923@qq．com2011-05-092011-08-10。———。。22．1。。。1。1。12．2λλλ2－4。1ftwjk2wjkwjkwjk－ujk3wjkft。2．3。。。。。4－7。fx=x|x|＞λ0|x|≤{λfx=sgnx|x|－λ|x|＞λ0|x|≤{λsgnxsgnn=1x＞0－1x{＜02．4。8－9。1DJλ=α2ln槡NαN。2λ=3αα。－3α3α6014203α3α10－11。3。3．1。。。。。fx=sgn|x|－pλ1+eq|x|－λ[]|x|≥λ0|x|＜{λλ。pqp∈01q≥0。|x|≥λ。pq。p=0qp∈01q=0p∈01q→∞p。pq。qp。pq。|x|≥λ、。3．2。。。λ=α2ln槡NjαNj。αNj。701No．12012。αα=median|fi|/0．槡6745fiα。αDα。λ=αD2ln槡Nj。αDα。44．1Matlab22b。p=0．9q=100022a128×128Camera2b0．20．012c2d2e2f2g2h。2、、。1lean。1。1Cameraleanlean7．61337．07547．53167．51277．57797．51247．53477．577965．466559．0078060．955960．741759．847462．005864．461864．681728．954223．2923．674423．942023．941923．888624．305727．4797801424．2pqpq。pq。q=0p0133p0．20．40．60．8122p0～1p、。p=0．9q0∞4。4q0105010050033q0～∞q。pq。3p∈01q=02p∈01q=0p=0．2p=0．4p=0．6p=0．8p=17．57297．56657．55737．54417．535261．628161．267260．918060．591160．286945．641845．173545．533943．729542．8004abcde4p∈0∞q=0．93p∈0∞q=0．9q=0q=10q=50q=100q=5007．54007．57767．57797．57797．577960．439161．999162．005762．005862．005843．280945．905445．910745．910945．91105。。。。1RafelCGonzalezRichardEWoods．DigitalImagepro-cessingM．2nded．PublishingHouseofElectronicsIn-dustry2007276－320．inChineseRafelCGonzalezRichardEWoods．M．2．2007276－320．2JixiangYan．InfradedImageSquenceEnhancementbasedonWaveletTransformD．Xi'anXi'anUniversityofElectronicScienceandTechnology．2008．21－19．inChinese．D．．2008．21－29．3SooChangKimTaeJinKang．AutomateddefectdetectionsystemusingwaveletpacketframeandGaussianmixturemodelC．OpticalSocietyofAmerica20062690－2701．4HaodeZengRenZhang．RestoringandenhancingofthinfogandclouddegradedimagesbasedonwaveletanalysisJ．JournalofPLAUniversityofScienceandTechnolo-gy．2005．76－80．inChinese901No．12012．J．．2005．76－80．5ChengZeng．Theresearchofimagede-noisinganden-hancementbasedonwavelettheoryJ．WuhanWuhanUniversityofTechnology200821－39．inChinese．D．200821－39．6JishaJohnMWilscy．EnhancementofweatherdegradedcolorimagesandvideosequencesusingwaveletfusionC．AdvancesinElectricalEngineeringandComputa-tionalScienceLectureNotesinElectricalEngineering200999－109．7JinjuLiZhifengMaQiongzhiWuetal．ImagedenoisebasedonthestationarywavelettranslationwithrotationJ．Laser＆Infrared201040111263－1268．inChinese．、J．201040111263－1268．8ChangjiangZhangXiaodongWangHaoranZhang．GlobalandlocalcontrastenhancementalgorithmforimageusingwaveletneuralnetworkandstationarywavelettransformJ．ChineseOpticalLetters2006636－639．9ZhiliangKang．Theresearchofimageenhancementalgo-rithmofinfradedimagebasedonwavelettransformD．ChengduUniversityofElectronicScienceandTechnologyofChina200525－40．inChinese．D．200525－40．10LiwenWang．SpecklesuppressionbasedonnonlinearmultiwavelettransformadapativethresholdvaluemethodJ．Laser＆Infrared2006362155－157．inChi-nese．J．2006362155－157．11LinZhangQianshengWangZhihuaXieetal．Multi-waveletadaptivethresholddenoisebasedongenerational-gorithmJ．Laser＆Infrared2008382186－190．inChinese．J．2008382櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆櫆186－190．01142