近年來人的身份辨識系統議題與研究越來越蓬勃發展，而其中以生物特徵且非接觸性的人臉影像作為系統識別身份的方法，最具親和與便利性。傳統人臉辨識所使用主成份分析（PCA）、特徵臉（EigenFace）及線性鑑別分析（LDA）等及其衍生的相關演算法，係直接使用整張影像成為向量信號，經過演算學習取得一投射向量擷取出特徵，其數學運算複雜而且繁複。本論文提出一個利用哈氏離散小波轉換（Haar Wavelet Transform），結合邊緣偵測索貝爾運算子（Sobel）所產生人臉灰階梯度變化，再擷取其正交投影信號作為特徵之人臉辨識系統。最後，採用奧立為特實驗室（Olivetti Research Lab, ORL）人臉資料庫，來驗證本文提出的人臉辨識系統之可行性與有效性。

Study of human face recognition system has grown vigorously for last decades. The biometrics of human face is an important feature in the identity authentication. The traditional principal component analysis (PCA), Eigenface, and linear decision algorithm (LDA) operate directly on a whole pattern represented as a vector and acquire a set of projection vectors to extract global features from given training patterns. The computation loads in these methods are heavy. In this thesis, a novel fusion algorithm of the wavelet transform, Sobel operator and orthogonal projection is proposed for human face recognition, where the Haar wavelet transform is utilized to reduce the dimension of images, the Sobel operator is adopted to extract facial features, and the orthogonal projection is presented to perform feature transformation and face recognition. The Olivetti Research Lab (ORL) face database is adopted to demonstrate the feasibility and effectiveness of the proposed face recognition system.

[1] AT&T Laboratories, Cambridge, U.K. The ORL Database of Faces [Online]. Available: http://www.orl.co.uk/facedatabase.html
[2] R. O. Duda, P. E. Hart and D. G. Stork, Pattern Classification. New York, NY: Wiley, 2000.
[3] 葉冠麟, Principal Component Analysis with Missing Data [Online]. Available: http://www.cmlab.csie.ntu.edu.tw/~cyy/learning/tutorials/PCAMissing Data.pdf
[4] J. R. Jang. Data Clustering and Pattern Recognition [Online]. Available: http://www.cs.nthu.edu.tw/~jang,
[5] A. Grossmann and J. Morlet, “Decomposition of Hardy functions into square integrable wavelets of constant shape,” SIAM J. Math., Vol. 15, pp. 723-736, 1984.
[6] S. Mallat, “A compact multiresolution representation: the wavelets model,” in Proc. IEEE Workshop Computer Vision, Miami, FL, Dec. 1987.
[7] Y. Meyer, “Ondelettes et Functions splines,” in Seminaire EDP, Ecole Polytechnique, Paris, France, Dec. 1986.
[8] G. Strang and T. Nguyen, Wavelets and Filter Banks. Wellesley, MA：Wellesley-Cambridge Press, 1996.
[9] P. S. Addison, Illustrated Wavelet Transform Handbook, Bristol： Institute of Physics Publishing, 2002.
[10] S. Mallat, “Multiresolution Approximation and Wavelet Orthonormal Bases of L2(R),” Trans. Amer. Math. Soc., Vol. 315, pp. 69-87, Sept. 1989.
[11] M. Vetterli, and C. Herley, “Wavelet and filter bank: Theory and Design,” IEEE Transactions on Signal Process, Vol. 40, No. 9, pp. 2207-2232, 1992.
[12] S. Du and R. Ward, “Wavelet-based illumination normalization for face recognition,” in Proc. IEEE Conf. Image Processing, Vol. 2, Sept. 2005, pp. 954-957.
[13] T. Acharya and A. K. Ray, Image processing: principles and applications. Hoboken, NJ： Wiley, 2005.
[14] 林宸生,應用於LCD上的定位研究. [Online]. Available: http://cslin.auto.fcu.edu.tw/oe/oe/lcd/index1.htm
[15] D.S. Moors and J.W. Yackel, “Large Sample Properties of Nearest Neighbor Density Function Estimators,” in Statistical Decision Theory and Related Topics, Academic Press, New York, pp. 269-279, 1976.
[16] B. V. Dasarathy, “Nearest Neighbor (NN) Norms: NN Pattern Classification Techniques,” IEEE Computer Society Press, Los Alamitos, CA, 1990.
[17] R. Chellappa, C. L. Wilson etc. “Human and Machine Recognition of faces: A survey”, Proc. the IEEE, Vol. 83, No. 5, pp 705-740, May, 1995,
[18] D. X. Shuicheng, Y. L. Zhang, M. Li, W. Ma, Z. Liu and H. Zhang, “Parallel Image Matrix Compression for Face Recognition,” in Proceedings of the 11th International Multimedia Modelling Conference, 2005, Jan. 2005, pp. 232–238.
[19] 鍾國亮, 影像處理與電腦視覺, 東華, 2006.
[20] 洪國勝、江國軍、龍國忠、洪月裡, C++ Builder 6 程式設計快樂上手, 旗標, 2002.
[21] 蕭富貴, Borland C++ Bulider 6 實用教學寶典, 台科大圖書, 2003.