人臉辨識系統主要可分為兩部份，一為人臉偵測，另一為人臉辨識。

人臉偵測方面，使用了影像前置處理和最大相似度的估計偵測單張影像中的人臉，在少數的限制情況下，我們的偵測方法可以克服如不同的距離、位置、人臉方向、臉部表情變化、遮蔽物(眼鏡)、和些微的光線改變。

在人臉辨識部份，使用卡式轉換和線性鑑別式分析來做為特徵萃取任務，這兩種方法做特徵萃取的過程中，特徵參數的取得是由原來資料和所選擇的特徵向量做內積運算，一般而言當資料庫變大時，辨識時間也會跟著增加，為了克服這個問題，本論文中使用硬極限卡式轉換來減少辨識時間。

The design of a face recognition system ( FRS ) can been separated into two major modules – face detection and face recognition.

In the face detection part, we combine image pre-processing techniques with maximum-likelihood estimation to detect the nearest frontal face in a single image. Under limited restrictions, our detection method overcomes some of the challenging tasks, such as variability in scale, location, orientation, facial expression, occlusion ( glasses ), and lighting change.

In the face recognition part, we use both Karhunen-Loeve transform and linear discrimant analysis ( LDA ) to perform feature extraction. In this feature extraction process, the features are calculated from the inner products of the original samples and the selected eigenvectors. In general, as the size of the face database is increased, the recognition time will be proportionally increased. To solve this problem, hard-limited Karhunen-Loeve transform ( HLKLT ) is applied to reduce the computation time in our FRS.

致謝辭………………………………………………………Ⅰ
論文提要……………………………………………………Ⅱ
中英文摘要………………………………………………ⅢⅣ
目錄…………………………………………………………Ⅴ
圖……………………………………………………………Ⅷ
表……………………………………………………………Ⅸ

第一章 緒論
1.1 研究動機…………………………………………1
1.2 人臉辨識系統流程………………………………4
1.3 論文架構…………………………………………6

第二章 相關文獻探討……………………………………… 7

第三章 人臉辨識
3.1 卡式轉換…………………………………………9
3.2 硬極限卡式轉換……………………………… 15
3.3 線性鑑別式分析……………………………… 18
3.4 分類法則……………………………………… 23

第四章 人臉偵測
4.1 偵測系統限制………………………………… 27
4.2 影像前置處理
4.2.1 灰階值正規化………………………… 29
4.2.2 直方圖均量化………………………… 30
4.3 眼睛特徵偵測與處理
4.3.1 二值化………………………………… 31
4.3.2 標記…………………………………… 32
4.3.3 找出可能的臉部候選區部…………… 35
4.4 人臉驗証
4.4.1 人臉模型訓練………………………… 37
4.4.2 驗証法則……………………………… 39

第五章 實驗結果
5.1 人臉資料庫…………………………………… 42
5.2 人臉偵測實驗結果…………………………… 43
5.3 人臉辨識實驗結果
5.3.1 實驗一：人數增加對辨識率之影響… 44
5.3.2 實驗二：訓練張數增加對辨識率之影響46
5.3.3 計算時間估計………………………… 48

第六章 結論與未來研究方向
6.1 結論……………………………………… 49
6.2 未來研究方向…………………………… 50

參考文獻…………………………………………………… 51

[1] Belhumeur, P.N. , Hespanha, J.P. , Kriegman, D.J. , “Eigenfaces vs. Fisherfaces: recognition using class specific linear projection” , IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 19 Issue: 7 ,July 1997
Page(s): 711 -720

[2] Brunelli, R. , Poggio, T. ,“Face recognition: features versus templates“, IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 15 Issue: 10 , Oct. 1993
Page(s): 1042 -1052

[3] C.C. Chen, C.T. Chen, and C.M. Tsai, Hard-limited Karhunen-Loeve transform for text-independent speaker recognition, Electronics Letters, 1997, 33, (24), pp. 2014-2016

[4] C.C. Chen and D.A. Landgrebe, “A spectral feature design system for the HIRIS/MODIS era”, IEEE Trans. Geosci. Remote Sens., 1989, 27, (6), Page(s): 681-686

[5] Chellappa, R. , Wilson, C.L. , Sirohey, S. ,“Human and machine recognition of faces: a survey“,Proceedings of the IEEE , Volume: 83 Issue: 5 , May 1995 Page(s): 705 -741

[6] Dana H. Ballard , Christopher M. Brown ,“ Computer vision ”, Prentice Hall, 1982

[7] Fukunaga, K. ,“Introduction to statistical pattern recognition“,Academic Press, San Diego, California, USA, 1990

[8] Lawrence, S. , Giles, C.L. , Ah Chung Tsoi; Back, A.D. , “ Face recognition: a convolutional neural-network approach“, IEEE Transactions on Neural Networks ,Volume: 8 Issue: 1 , Jan. 1997 Page(s): 98 -113


[9] Martinez, A.M. ,“ Recognizing imprecisely localized,partially occluded and expression variant faces from a single sample per class“,IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 24 Issue: 6 , June 2002 Page(s): 748 -763

[10] Martinez, A.M. , Kak, A.C. ,“PCA versus LDA”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume: 23 Issue: 2 , Feb. 2001 ,Page(s): 228 -233

[11] Ming-Hsuan Yang , Kriegman, D.J. , Ahuja, N. , “Detecting faces in images: a survey ”, IEEE Transactions on Pattern Analysis and Machine Intelligence,, Volume: 24 Issue: 1 , Jan. 2002 Page(s): 34 -58

[12] Moghaddam, B. , Pentland, A. ,“ Probabilistic visual learning for object representation “, IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 19 Issue: 7 , July 1997 Page(s): 696 -710

[13] Rafael C. Gonzalez , Richard E. Woods ,“Digital image processing“, Prentice Hall, New Jersey, 2002

[14] Rein-Lien Hsu , Abdel-Mottaleb, M. , Jain, A.K. ,“Face detection in color images ”, International Conference on Image Processing, 2001. Proceedings. Volume: 1 , 7-10 Oct. 2001 Page(s): 1046 -1049 vol.1

[15] Rowley, H.A. , Baluja, S. , Kanade, T. ,“ Neural network-based face detection“, IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 20 Issue: 1 , Jan. 1998 Page(s): 23 -38

[16] Srisuk, S. , Kurutach, W. ,“A new robust face detection in color images “, Fifth IEEE International Conference on Automatic Face and Gesture Recognition, 2002. Proceedings., 20-21 May 2002 Page(s): 291 -296


[17] Swets, D.L. , Weng, J.J. ,“Using discriminant eigenfeatures for image retrieval ”, IEEE Transactions on Pattern Analysis and Machine Intelligence,Volume: 18 Issue: 8 , Aug. 1996 Page(s): 831 -836

[18] Theodoridis, S. , Koutroumbas, K. ,“Pattern recognition“, Academic Press, San Diego, 1999

[19] Turk, M.A. , Pentland, A.P. ,“ Face recognition using eigenfaces“, IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1991. Proceedings CVPR '91., 3-6 June 1991 Page(s): 586 -591

[20] Turk, M.A. , Pentland, A.P. ,“Eigenfaces for recognition“, Journal of Cognitive Neuroscience Volume 3,Number 1,1991 Massacbusetts Institute of Technology Page(s): 71-86

[21] Wiskott, L. , Fellous, J.-M. , Kruger, N. , von der Malsburg, C. ,“Face recognition by elastic bunch graph matching“,
International Conference on Image Processing, 1997. Proceedings.,Volume: 1 , 26-29 Oct. 1997 Page(s): 129 -132 vol.1

[22] Yanjiang Wang , Baozong Yuan ,“Segmentation method for face detection in complex background”, Electronics Letters , Volume: 36 Issue: 3 , 3 Feb. 2000 Page(s): 213 -214