搜索
西华大学毕业设计说明书摘要人脸识别（Facerecognition）是指基于人的脸部特征信息进而实现对身份识别的一种生物识别技术，它是现在生物模式识别领域里非常重要和实用的人工智能研究课题。相比于指纹、虹膜识别来说，人脸识别是非接触式，运用起来会更加直观、方便。近年来，人脸识别技术引起了越来越多的研究人员的重视。人脸识别技术的研究取得了很大程度上的进展，它现在广泛的应用于日常生活的各种领域，如网络的安全、公安、电子商务、金融、考勤和物业管理等。本文针对人脸预处理、特征提取和分类识别方面进行了研究和分析。首先，分析了光线、姿态、表情、角度等因素对图像质量的影响，明确了人脸预处理的必要性并介绍常用的预处理方法。然后，介绍和分析Gabor小波变换和主成分分析方法（PCA），用于人脸特征提取。本文分别采用最近邻分类器和SVM分类器识别人脸特征。并使用AR人脸库进行实验，分别通过不同的实验结果来分析预处理的作用、Gabor小波和PCA的性能，以及最近邻和支持向量机两种分类器的性能比较。【关键词】人脸识别预处理Gabor小波支持向量机西华大学毕业设计说明书AbstractFacerecognitionisakindofbiometricidentificationtechnologybasedonfacialfeaturesofpeople'sinformation,isaresearchsubjectofartificialintelligenceinthefieldofbiologicalidentificationisveryimportantandpractical.Comparedtothefingerprintandirisrecognition,facerecognitionisnon-contact,usemoreintuitiveandconvenient.Inrecentyears,facerecognitiontechnologyhasattractedmoreandmoreattentions,progresshasbeenmadelargelyoffacerecognition,itisnowwidelyusedinvariousfieldsofdailylife,suchasnetworksecurity,publicsecurity,electroniccommerce,finance,attendanceandpropertymanagementetc.Inthispaper,theaspectsoffacepreprocessing,featureextractionandclassificationandrecognitionarestudiedandanalyzed.Firstly,theinfluenceofthefactorssuchasray,attitude,expressionandangleontheimagequalityisanalyzed,andthenecessityoffacepreprocessingisclarified,andthecommonpreprocessingmethodisintroduced.Then,theGaborwavelettransformandprincipalcomponentanalysis(PCA)areintroducedandanalyzed,whichisusedforthefeatureextractionofhumanface.Inthispaper,thefeatureoffaceischaracterizedbythenearestneighborclassifierandSVMclassifier.AndexperimentswereconductedusingtheARfacedatabase,respectivelybydifferentexperimentalresultstoanalyzetheperformanceofpretreatment,GaborwaveletandPCA,andthenearestneighborandsupportvectormachineclassifierperformancecomparison.【KeyWords】FacerecognitionPreprocessingGaborwaveletSupportvectormachineI西华大学毕业设计说明书目录前言................................................................................................................................11人脸识别介绍.........................................................................................................21.1人脸识别发展历史......................................................................................21.2人脸识别系统构成......................................................................................21.2.1人脸图像采集及检测........................................................................21.2.2人脸图像预处理................................................................................31.2.3人脸图像特征提取............................................................................31.2.4人脸图像匹配与识别........................................................................31.3人脸识别常用方法......................................................................................32基于人脸的预处理.................................................................................................52.1直方图均衡化..............................................................................................52.1.1基本原理............................................................................................52.1.2实现方法............................................................................................63人脸特征提取与识别.............................................................................................93.1Gabor小波变换基本原理...........................................................................93.2主成分分析法基本原理..............................................................................93.3分类方法....................................................................................................103.3.1最近邻法（NN）基本原理............................................................103.3.2支持向量机（SVM）基本原理......................................................124系统设计与实验结果...........................................................................................154.1设计目的....................................................................................................154.2功能实现....................................................................................................154.2.1Gabor小波变换..............................................................................154.2.2主成分分析法..................................................................................18II西华大学毕业设计说明书4.2.3最近邻法..........................................................................................184.2.4支持向量机......................................................................................194.3工作流程图................................................................................................204.4不同实验结果分析....................................................................................214.4.1不进行预处理..................................................................................224.4.2进行预处理......................................................................................234.5存在的问题和不足....................................................................................245Android平台人脸辨识实现以及开发环境...............