简介:摘要人脸识别是模式识别以及图像处理研究的重要内容和热点之一,也是生物特征识别技术中的一个非常活跃的课题。人脸识别技术具有实时、准确和非接触等优势,因而较容易被用户接受和认可,目前已经在出入管理、门禁考勤等系统中有着广泛的应用。本文主要研究基于Gabor小波变换与协同表示的人脸识别算法,就其中的特征提取等相关问题进行了深入探讨。首先用Gabor小波对人脸图像进行特征提取,由于变换后的特征维数较高,所以要对变换特征进行降维,本文结合协同表示的方法提出了一种新的基于Gabor特征与协同表示的人脸识别算法,实验结果表明该方法对于人脸图像的光照、表情和姿态等变化具备较强鲁棒性,算法运行速度较快。
简介:这篇文章重点研究改进的Gabor小波(improvedGaborwavelet,IGW)变)并讨论了它在地震信号处理和解释中的应用ThispaperintroducesanimprovedGaborwaveletanditscompletetransform,andmainlyanalysestheirpropertiesanddiscussesapplicationsofthesepropertiesinseismicsignalprocessandinterpretation。改进的Gabor小波变换具有以下特性:1)IGWT把时域信号映射到时间一频率域,而传统Gabor小波变换把时间信号映射到时间一尺度域;2)IGWT可用于信号分频,通过固定变换的主频参数dominantfrequency,并变换能提取相应的子带信号,且其主频部分的信息与原信号相应频率部分的信息一致,通过调节变换的分辨率因子,变换能有效控制子带信号的带宽;3)用IGWT和IGWIT构建的滤波器有良好的时一频局部性,在指定时一频范围内能实现针对性滤波。文章用仿真实验和实际用例验证IGWT的这些特性,并在提高地震信号分辨率、地震信号分频和识别小断层等地震信号处理和解释等方面的应用中取得良好效果。
简介:Inthisworktwoaspectsoftheoryofframesarepresented:asidenecessaryconditiononirregularwaveletframesisobtained,anotherperturbationofwaveletandGaborframesisconsidered.Specifically,wepresenttheresultsobtainedonframestabilitywhenonedisturbsthemotherofwaveletframe,ortheparameterofdilatation,andinGaborframeswhenthegeneratingfunctionortheparameteroftranslationareperturbed.Inallcasesweworkwithoutdemandingcompactnessofthesupport,neitheronthegeneratingfunction,noronitsFouriertransform.
简介:给定的L2(R)和一个有限序列{(r,r)}rR+XR由不同的点组成,相应小浪系统是函数$\left\的集合{{\frac{1}{{a_\gamma^{1/2}}}\phi(\frac{x}{{a_\gamma}}-\lambda_\gamma)\gamma\varepsilonr}\right\}$。我们为功能L2(R)的一个稠密的集合证明那相应于任何选择的小浪系统{(r,r)}r是线性地独立的,并且我们导出清楚的估计为相应更低(框架)跳。特别地,这为multiresolution分析在理论把限制放在可伸缩的功能的选择上。我们也为Gabor系统$\left\为更低的界限获得估计{{e^{2rie_{\gammax}}g(x-\lambda_\gamma)}\right\}为在L2(R)的一个稠密的子集的函数g的\gamma\varepsilonr$。
简介:ConventionalGaborrepresentationanditsextractedfeaturesoftenyieldafairlypoorperformanceinextractingtheinvariancefeaturesofobjects.Toaddressthisissue,aglobalGaborrepresentationmethodforraisedcharacterspressedonlabelisproposedinthispaper,wheretherepresentationonlyrequiresfewsummationsontheconventionalGaborfilterresponses.Featuresarethenextractedfromthesenewrepresentationstoconstructtheinvariantfeatures.ExperimentalresultsclearlyshowthattheobtainedglobalGaborfeaturesprovidegoodperformanceinrotation,translation,andscaleinvariance.Also,theyareinsensitivetoilluminationconditionsandnoisechanges.ItisprovedthatGaborfilterscanbereliablyusedinlow-levelfeatureextractioninimageprocessingandtheglobalGaborfeaturescanbeusedtoconstructrobustinvariantrecognitionsystem.
简介:Recently,ShiXianliangandHuLanpublishedthemethodofconcentrationfactorsfordeterminationofjumpsoffunctionsviaMCMconjugatewavelets.Usually,itisdiculttocalculatetheHilberttransformofgeneralwindowfunctions.TheaimofthispaperistodiscussdeterminationofjumpsforfunctionsbasedonderivativeGaborseries.Theresultswillsimplifythecalculationofjumpvalues.
简介:Gabor特征被显示了为手掌静脉识别有效。这份报纸论述一个新奇特征表示方法,实现本地Gabor直方图(FLGH)的熔化,以便改进手掌静脉识别系统的精确性。称为本地Gabor主要差别模式(LGPDP)的一个新本地描述符用本地最大的差别(LMD)编码Gabor大小操作员。相应Gabor阶段模式被独占的本地Gabor编码或(XOR)模式(LGXP)。Fishers线性判别式(FLD)方法然后被实现减少特征表示的维数。低维的Gabor大小和阶段特征向量最后被熔化提高精确性。从自动化的研究所的试验性的结果,建议FLGH方法更好完成的中国科学院(CASIA)数据库表演由利用分数级的熔化的性能。相等的错误率(无论何时)是0.08%,它超过另外的常规手掌静脉识别方法(从2.87%~0.16%的无论何时范围),例如,拉普拉斯算符手掌,琐事特征,麻袋布阶段,Eigenvein,本地不变的特征,相互的前景本地人二进制代码模式(LBP),并且多采样特征熔化方法。
简介:Inthispaper,weproposeasparseovercompleteimageapproximationmethodbasedontheideasofovercompletelog-Gaborwavelet,meanshiftandenergyconcentration.Theproposedapproximationmethodselectsthenecessarywaveletcoefficientswithameanshiftbasedalgorithm,andconcentratesenergyontheselectedcoefficients.Itcansparselyapproximatetheoriginalimage,andconvergesfasterthantheexistinglocalcompetitionbasedmethod.Then,weproposeanewcompressionschemebasedontheaboveapproximationmethod.TheschemehascompressionperformancesimilartoJPEG2000.TheimagesdecodedwiththeproposedcompressionschemeappearmorepleasanttothehumaneyesthanthosewithJPEG2000.
简介:Fundusdiagnosisisanimportantpartofthewholebodyexaminationthatmayproviderichclinicalinformationtodoctorsfordiagnosticreference.Manualfundusvesselextractionishelpfultoquantitativemeasurementofdiseasesbutobviouslyitisatoughworkforphysicians.ThispaperpresentsanautomaticmethodbyusingGaborfilterbanktoextractthearteryandveinseparatelyintheocularfundusimages.Afterpreprocessingstepsthatincludegray-scaletransform,grayvalueinversionandcontrastenhancement,theGaborfilterbankisappliedtotheextractionofthearteryandveinintheocularfundusimages.Finallythesetwodifferentwidthtypesofvesselsareselectedbypost-processingmethodssuchaslabeling,corrosion,binarization,etc.Evaluationresultsshowanaccuraterateof90%inveinand82%inarteryfrom20cases,thatindicatestheeffectivenessofourproposedsegmentationmethod.
简介:Alocalizedparametrictime-shearedGaboratomisderivedbyconvolvingalinearfrequencymodulatedfactor,modulatinginfrequencyandtranslatingintimetoadilatedGaussianfunction,whichisthegeneralizationofGaboratomandismoredelicateformatchingmostofthesignalsencounteredinpractice,especiallyforthosehavingfrequencydispersioncharacteristics.Thetime-frequencydistributionofthisatomconcentratesinitstimecenterandfrequencycenteralongenergycurve,withthecurvebeingobliquetoacertainextentalongthetimeaxis.Anovelparametricadaptivetime-frequencydistributionbasedonasetofthederivedatomsisthenproposedusingaadaptivesignalsubspacedecompositionmethodinfrequencydomain,whichisnon-negativetime-frequencyenergydistributionandfreeofcross-terminterferenceformulticomponentsignals.Theresultsofnumericalsimulationmanifesttheeffectivenessoftheapproachintime-frequencyrepresentationandsignalde-noisingprocessing.