基于LBP-KNN和CNN-SVM的人臉識(shí)別算法
發(fā)布時(shí)間:2021-06-22 00:27
人臉在我們社會(huì)交往中扮演著重要的角色,傳遞著我們自身信息。生物識(shí)別密碼技術(shù)是一種非常關(guān)鍵的安全技術(shù),因其有著廣泛的應(yīng)用前景,在過去的幾年里一直受到業(yè)內(nèi)廣泛的關(guān)注。人的面部表情有很多的變化(如:臉部老化、面部表情、明亮程度、不標(biāo)準(zhǔn)的姿勢(shì)等),這些變化會(huì)導(dǎo)致臉部識(shí)別信息不準(zhǔn)確,辨認(rèn)身份能力較差。雖然人臉識(shí)別的技術(shù)上,已經(jīng)有了很大的進(jìn)展,同時(shí)也顯示了非常精確的結(jié)果,但是在實(shí)際應(yīng)用中,年齡不變的人臉識(shí)別仍然是系統(tǒng)應(yīng)用中一個(gè)非常重要的挑戰(zhàn)。我們研究的目的是提供一種解決臉部識(shí)別問題的方法,這些問題收很多參數(shù)變化的影響,如姿勢(shì)、明亮程度、年齡不變和面部表達(dá)等。為了解決這些問題,下一節(jié)將詳細(xì)闡述不同的算法,來證明所提出模型的有效性。為了證明在姿態(tài)變化、明亮程度和表達(dá)方面獲取結(jié)果的可靠性,我們結(jié)合了兩種算法:(a)魯棒性local binary pattern(LBP),用于面部特征提取;(b)k-nearest neighbor(K-NN)進(jìn)行圖像分類。我們的實(shí)驗(yàn)已經(jīng)在CMU PIE(Carnegie Mellon University Pose,Illumination,and Expression...
【文章來源】:杭州電子科技大學(xué)浙江省
【文章頁(yè)數(shù)】:72 頁(yè)
【學(xué)位級(jí)別】:碩士
【文章目錄】:
摘要
ABSTRACT
Chapter1 Introduction
1.1 Problem Definition,Motivation and Objectives
1.2 Ethic and Society Implications
1.3 Literature Survey
1.3.1 Pose,Illumination and Expression Face Recognition
1.3.2 Age Invariant Face Recognition
1.4 Resume
Chapter2 Face Recognition Fundamental
2.1 History of Face Recognition
2.2 Face Recognition System
2.3 Different Face recognition challenges
2.3.1 Pose Variation issues
2.3.2 Illumination Variation issues
2.3.3 Expression Variation issues
2.3.4 Age Invariant issues
2.3.5 Other related issues:Plastic/Cosmetic Surgery and Makeup
2.4 Resume
Chapter3 Face Recognition using Local Binary Pattern And K Nearest Neighbor
3.1 Local Binary Pattern(LBP)
3.2 Using K-Nearest Neighbor(KNN)to Classify a Face Image
3.3 Gaussian Filter
3.4 Proposed Face Recognition System
3.5 Test Results for CMU PIE and LFW
Chapter4 Face Recognition using Convolutional Neural Network and Support Vector Machine
4.1 Convolutional Neural Network(CNN)
4.2 Support Vector Machine(SVM)
4.3 Proposed Aging Face Recognition
4.4 Test Results for MORPH Album2 and FG-Net
Chapter5 Conclusion
Acknowledgement
References
Appendix
本文編號(hào):3241736
【文章來源】:杭州電子科技大學(xué)浙江省
【文章頁(yè)數(shù)】:72 頁(yè)
【學(xué)位級(jí)別】:碩士
【文章目錄】:
摘要
ABSTRACT
Chapter1 Introduction
1.1 Problem Definition,Motivation and Objectives
1.2 Ethic and Society Implications
1.3 Literature Survey
1.3.1 Pose,Illumination and Expression Face Recognition
1.3.2 Age Invariant Face Recognition
1.4 Resume
Chapter2 Face Recognition Fundamental
2.1 History of Face Recognition
2.2 Face Recognition System
2.3 Different Face recognition challenges
2.3.1 Pose Variation issues
2.3.2 Illumination Variation issues
2.3.3 Expression Variation issues
2.3.4 Age Invariant issues
2.3.5 Other related issues:Plastic/Cosmetic Surgery and Makeup
2.4 Resume
Chapter3 Face Recognition using Local Binary Pattern And K Nearest Neighbor
3.1 Local Binary Pattern(LBP)
3.2 Using K-Nearest Neighbor(KNN)to Classify a Face Image
3.3 Gaussian Filter
3.4 Proposed Face Recognition System
3.5 Test Results for CMU PIE and LFW
Chapter4 Face Recognition using Convolutional Neural Network and Support Vector Machine
4.1 Convolutional Neural Network(CNN)
4.2 Support Vector Machine(SVM)
4.3 Proposed Aging Face Recognition
4.4 Test Results for MORPH Album2 and FG-Net
Chapter5 Conclusion
Acknowledgement
References
Appendix
本文編號(hào):3241736
本文鏈接:http://sikaile.net/kejilunwen/zidonghuakongzhilunwen/3241736.html
最近更新
教材專著