thereportofthefacerecognition_英语考试_外语学习_教育专区

《thereportofthefacerecognition_英语考试_外语学习_教育专区》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、IntroductionBackgroundEigenfacesisacrucialconceptconcerningthecomputervisionofhumanityfacerecognition.AccordingtotheillustrationofMatthewandAlexcite{l},theeigenfacesactualareasetofeigenvectorswhichusedtorepresentimageswiththecorrespondingappropriatew

2、eightvalues.ThismechanismisextremelysimilarwithutilizingtheCartesianCoordinationSystemtodescribepointsinthethree-dimensionspace・Thelatterutilizesthex,y,andzaxestodeterminethepositionofapointandtheformerutilizestheeigenfacestodetermineanimage・Inthispro

3、cess,asetofeigenfacescanbecalledasetofdescribebasisforrecognizingimages・IntroductionInordertounderstandpreliminarilythebasicprincipleofhumanfacerecognitionbycomputerandtheroleoftheeigenfacesisplayedinthisprocess,thisexperimentisdividedintosixsectionsa

4、ndthesoftwareemph{matlab}isutilized・Beforethelab,theresomerequisiteoperationsarenecessarytobemanipulated.Todownloadthefilenamedemph{data_for_labC.mat}intotheemph{Matlab}andthentoinspectthevariablesemph{eigenfaces_blk}andemph{employees_DB}.Thefo

5、rmercontains101parametersandthefirst100ofthemaretheeigenfacesduetotheirface-likeappearaneeiftheyaredisplayedasimages.Thelastparameterofthevariableemph{eigenfaces_blk}isactuallyredundantinthislab,whichwillbeexplainednextinthefollowingpaper.Thevariable

6、emph{employees_DB}isastructthatcontaintheIDnumbersofthe100employeesandthecorrespondingweightsdescribingthefacesoftheseemployees・Thetaskofeachsectionislistedasbelow.:LTaskl:Toverifythattheseeigenfaceinthevariableemph{eigenfaces_blk}areorthogonal.2.Ta

7、sk2:Withthegivenegienfaces,theweightsoftheinputimageisfoundbyproject!ngitontheentirecollect!onoftheegienfaces(thetrainingfaces).Sincetheinputimageprojectingoneacheigenfacewillgenerateaweightvalue,therewillgainasetofweightsthatisthesamesizewiththenumbe

8、rofeigenfaces.Thereneedstowriteaemph{Matlab}functiontogaintheweightsofaface・Thedeclarationofthefunctionshouldbelike:$Function[weight_of_face,meanof_face]=get_^ace_weights(im,eigenfaces_blk)$Whereemph{mean_of_face}istheaverageintensityo