Human Face Recognition using Line Features

Human Face Recognition using Line Features

ID:40555911

大小:155.50 KB

页数:8页

时间:2019-08-04

Human Face Recognition using Line Features_第1页
Human Face Recognition using Line Features_第2页
Human Face Recognition using Line Features_第3页
Human Face Recognition using Line Features_第4页
Human Face Recognition using Line Features_第5页
资源描述:

《Human Face Recognition using Line Features》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库

1、HumanFaceRecognitionusingLineFeaturesMrinalKantiBhowmik1,DebotoshBhattacharjee2,MitaNasipuri2,DipakKumarBasu2,MahantapasKundu21DepartmentofComputerScienceandEngineering,TripuraUniversity,Suryamaninagar-799130,Tripura,India2DepartmentofComputerScienceandEngineering,J

2、adavpurUniversity,Kolkata-700032,IndiaABSTRACTInthisworkweinvestigateanovelapproachtohandlethechallengesoffacerecognition,whichincludesrotation,scale,occlusion,illuminationetc.Here,wehaveusedthermalfaceimagesasthosearecapabletominimizetheaffectofilluminationchangesa

3、ndocclusionduetomoustache,beards,adornmentsetc.Theproposedapproachregistersthetrainingandtestingthermalfaceimagesinpolarcoordinate,whichiscapabletohandlecomplicaciesintroducedbyscalingandrotation.Linefeaturesareextractedfromthermalpolarimagesandfeaturevectorsarecons

4、tructedusingtheseline.Featurevectorsthusobtainedpassesthroughprincipalcomponentanalysis(PCA)forthedimensionalityreductionoffeaturevectors.Finally,theimagesprojectedintoeigenspaceareclassifiedusingamulti-layerperceptron.IntheexperimentswehaveusedObjectTrackingandClas

5、sificationBeyondVisibleSpectrum(OTCBVS)database.Experimentalresultsshowthattheproposedapproachsignificantlyimprovestheverificationandidentificationperformanceandthesuccessrateis99.25%.Keywords:Facerecognition,Thermalfaceimages,Polartransformation,Principalcomponenta

6、nalysis,Linefeatures,Multilayerneuralnetwork,Backpropagationlearning,Classification1.INTRODUCTIONTobesafeandsecureinthepresentworlditisrequiredtoauthenticateanauthorizedperson,tokeepsuspectsunderconstantsurveillance,andalsotoinvestigateaboutcriminals,missingandtrace

7、lessindividualswithinareasonablecoststructureisamajorproblem.Asecuritysystembasedonhumanfaceisnon-intrusiveinnature,i.e.withoutanyconsentfromthepersonconcerned,he/shecanbeinvestigated,whichisnottrueforothertypesofbiometricsignatureslike,finger-print,handgesture,reti

8、naetc.Humanfaceimageanalysislacksrobustnessduetodifferentvariationandthesourcesarebeard,moustache,spectacles,hair,adornmentsetc.Thermalfac

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。