近紅外人臉圖像的分類研究

1.24萬字
我自己的畢業(yè)論文，原創(chuàng)的，已經通過校內系統檢測，僅在本站獨家出售，重復率低，大家放心下載使用

摘要 近年來人們越來越多的關注生物特征識別，而在這些生物特征識別方法中，人臉識別具有方便，經濟而準確的特點，可以廣泛應用于高安全性部門的警戒、入口控制、人機交互、計算機保密、公共安全等方面。在人臉識別中，人臉外觀會受到光照、姿態(tài)、表情變化的影響，所以人臉識別系統要適應各種環(huán)境，此外化妝、照片欺詐也是人臉識別亟待解決的問題。本文主要研究基于二維線性判別分析的人臉識別。
本文首先介紹人臉識別的研究意義、識別方法、研究中的問題和難點以及人臉識別的現狀和發(fā)展前景。研究以近紅外圖像為基礎的人臉識別方法，并且也了解近紅外圖像人臉識別的特點和價值，它的缺點和可提高識別性能的研究方向也是要了解的。根據以上內容，本文著重研究二維線性判別分析（2DLDA）方法。2DLDA算法識別率高、處理速度快，可以對人臉圖像數據降維，將人臉圖像測試樣本變換為二維矩陣并進行列/行方向的二維線性判別分析。這是2DLDA算法的優(yōu)勢。
在Matlab上對基于二維線性判別分析的人臉識別算法進行編程實現。
關鍵詞 近紅外 人臉識別 二維線性判別


Classification of near infrared face image

Abstract In recent years, more and more people were concerned about biometric identification, and in these biometric identification methods, face recognition with the characteristic of convenient, economical and accurate, can be widely used in high security alert authorities, access control, human interactive, computer secrecy, public safety and other aspects. In face recognition,light, gesture, facial expression affected face appearance, so the face recognition system have to adapt to the environment. In addition to makeup, photo recognition fraud were also serious problems. This paper studied the research of face recognition based on two-dimensional linear discriminant analysis.
This paper introduced the significance of face recognition, identification methods, research problems and difficulties as well as research status of face recognition and development prospects. The paper discussed the characteristics and values of face recognition based on the near-infrared image ​, comprehended the face recognition methods based on near-infrared image, meanwhile discussed the shortcomings of face recognition research based on the near-infrared image and improve recognition performance. On this basis, the paper focused on the two-dimensional linear discriminant analysis (TDLDA) method. 2DLDA algorithm had higher recognition rate, fast processing speed, can reduce the dimensionality of face image data, the face image test samples were converted into a two-dimensional matrix and thus made two-dimensional linear discriminant analysis of row/column direction. This is the advantages of 2DLDA algorithm.
A program about the algorithm of face recognition based on 2DLDA is written on MATLAB.
Key words Near-infrared facial recognition 2DLDA

目錄
第一章 緒論…………………………………………………………………………........1
1.1 研究人臉識別的意義……………………………………………..............................1
1.2 人臉識別的研究內容……………………………………………..............................1
1.3 人臉識別研究現狀及難點………………………………………..............................1
1.4 人臉識別系統……………………………………………………..............................2
1.5 人臉識別的發(fā)展趨勢及應用領域………………………………..............................2
1.6 本文主要工作及結構安排……………………………………………………...….3
第二章 近紅外人臉識別及二維線性判別分析………..………............................ 4
2.1 近紅外人臉識別……………………………………………………………...….....4
2.1.1 近紅外人臉識別的意義……………………………………………………...……4
2.1.2 近紅外圖像…………………………………………………………………...……4
2.2 人臉圖片庫……………………………………………………………………...….5
2.3 線性判別分析…………………………................ ………………...........................6
2.3.1 LDA與2DLDA………………………………………………………………........6
2.3.2 線性判別分析的缺點與改進………………………………………………….......6
第三章 基于二維線性判別分析的人臉識別………………………………….…...8
3.1 線性判別分析的原理…………………………………………………………..........8
3.2 線性判別法……………………………………………………………………...….9
3.3 2DPCA原理……………………………………………………………...………….10
3.4 2DLDA原理……………………………………………………………...………….10
3.5 2DLDA在人臉識別中的應用…………………………………….…………...……12
3.5.1 特征提取……………………………………………………………………….....12
3.5.2 分類…………………………………………………………………………..…...12
3.6 算法評價標準…………………………………………………………………….....13
第四章 MATLAB編程實現及運行結果分析……………………………………... 14
4.1 MATLAB相關函數簡介………..