运动和散焦模糊图像的复原方法及其应用研究

3710Vo.l37,No.10200710LASER&INFRAREDOctober,2007:10015078(2007)10112003王枚1,3,潘国华2,王国宏1,尤晶晶3(1.,264001;2.,264000;3.,264000):模糊图像复原是从退化的观测图像中重建原始图像的过程,是进行图像处理和模式识别等研究的基础运动模糊和散焦模糊是两类常见的模糊方式通过对这两类模糊图像复原过程的研究,分析了模糊图像复原过程中滤波器的选择点扩散函数的确定以及边缘噪声的处理等关键技术,提出了图像复原的可行方案计算机仿真试验结果表明,该方法是有效地和可行的:模糊图像;图像复原;点扩散函数;滤波器;傅里叶变换:TP391.41:ATheRestorationMethodofMotionblurredandDefocusblurredImagesandItsApplicationWANGMei1,3,PANGuohua2,WANGGuohong1,YOUJingjing3(1.DepartmentofElectronicandInformationEngineering,NavalAeronauticalEngineeringInstitute,Yantai264001,China;2.YantaiPublicSecurityBureau,Yantai264000;3.DepartmentofComputerandInformationEngineering,YantaiVocationalCollege,Yantai264000,China)Abstract:Thepurposeofblurredimagesrestorationisthattheoriginalimageisrestructuredfromtheblurredimages.Theblurredimagerestorationisthebasetechnologyforimagesprocessingandpatternrecognition.Themotionblurredanddefocusblurredaretwokindsofblurredstyles.Theeffectualapproachispresentedbytheresearchoffilter,spreadfunctionandedgenoiseintheblurredimagesrestorationprocessing.Thecomputerrestorationexperimentsonactualblurredimagesshowthatthemethodiseffectiveandpractica.lKeywords:blurredimage;imagerestoration;pointspreadfunction;filter;Fouriertransform1,,,:LucyRichardson[1-8],1,,;,12,,:(No.60541001);(No.200443):(1968-),,,,Emai:[e-mail]:20070423,,,;,,,,,(InverseFiltering)(WienerFiltering),,2.1逆滤波和维纳滤波,g(x,y)f(x,y)PSFh(x,y),:g(x,y)=-f(,)h(x-,y-)dxdy+n(x,y)(1),n(x,y),(1):G(u,v)=F(u,v)H(u,v)+N(u,v)(2),G(u,v),H(u,v),N(u,v)g(x,y),h(x,y),n(x,y),H(u,v)(MTF),F(u,v),F^(u,v)=G(u,v)H(u,v)(3)(3),,:E(u,v)=F^(u,v)-F(u,v)=N(u,v)H(u,v)(4),H(u,v),,;,N(u,v),H(u,v),,E(u,v),,H(u,v),,:F^(u,v)=H*(u,v)H(u,v)H*(u,v)+rG(u,v)(5),H*(u,v)H(u,v);r(5),,,r0.0001~0.01[4],r0.01,(Ghost);r0.0001,(Ring)r,H(u,v):Hwiener(u,v)=H*(u,v)|H(u,v)|2+r(6)2.2点扩散函数的确定,PSF,PSFH(u,v)[9],:h(i,j)=1/(R2)i2+j2R0(7),R:h(i)=1/L-L/2iL/20(8),L,[10],(6)2.3边缘误差与退化边缘,,,,,,,,0,,,,a!b!!!c!d,a!bb!b!!!!!!c!dd!dc!dd!d!!!!!!c!dd!d(9)1121No.102007a!bb!a!!!!!!c!dd!cc!dd!d!!!!!!a!bb!a(10),,,,,3:r;,;,:STEP1:ISTEP2:PSF,;,[7],(7)(8)PSFSTEP3:PSF,(6)STEP4:I(9),(64∀64,128∀128,256∀256)I#STEP5:I#,STEP6:I#,STEP7:,()STEP8:,STEP9:(),I(64∀64)23(a)(b)2Fig.2therestorationresultforsmallregionimagesmotionblurredwithnoise(a)10(b)3Fig.3therestorationresultforsmallregionimagesdefocusblurredwithnoise(288∀768),4,4Fig.4therestorationresultfortheactualvehicleimagesmotionblurredanddefocusbturredwithnoise,,,,,,PSF,[11][12],(7)(8)RL4(下转第1126页)1122372.3.2基于数学形态学的有效区域提取,,,4:,,,,(),(),,3(b),,3AFRS,,,:[1]RossA,JainAK,ReismanJ.Ahybridfingerprintmatcher[J].PattemRecognition,2003,36(7):1661-1673.[2]TicoM,KuosmanenP.Fingerprintmatchingusinganorientationbasedminutiadescriptor[J].IEEETransactionsonPatternAnalysisandMachineIntelligence,2003,25(8):1009-1014.[3]LHong,AKJain,RBolle.Fingerprintenhancement[J].Proc.FirstIEEEWACV,1996:202-207,Sarasota,Fla.[4]AmengnalJC,JuanA,PerezJC,eta.lRealtimeminutiaeextractioninfingerprintimages[C]//Proc.ofthe6thInt.Con:fOnImageproceedingandApplications(IPA97)[EB/OL][5]JiangXY,MojonD.Adaptivelocalthresholdingbyverification2basedmultithresholdprobingwithapplicationtovesseldetectioninretinalimages[J].IEEETransactionsonPatternAnalysisandMachineIntelligence,2003,25(1):131-137.[6].[M].:,2000.(上接第1122页),,r,;;,,:[1],.[J].,2005.8:23-25.[2]MarkRB,AggelosK.Digitalimagerestoratmn[J].IEEESignalProcessingMagazine,1997,14(2):24-41.[3],,,.[J].,2005,33(9):1710-1713.[4]RafaelCG,RichardEW,StevenLE.DigitalimageprocessingusingMATLAB[M].Beijing:PublishingHouseofElectronicsIndustry,2005,7:141-178.(inChinese)[5],.[J].,2005.33(7):1288-1290.[6],,.[J].,2005,13(6):735-740.[7],,.[J].,2005,15(6):28-32.[8],.[J].∃,2005,16(9):1107-1111.[9].[M].:,2002:34-36.[10].[J].(),2005,32(4):376-378.[11],.[J].,2006.27(5):514-516.[12],.[J].,2005,25(6):1316-1317.112637