臺灣博碩士論文加值系統

主值分析法(principal component analysis, PCA)和獨立值分析法(independent component analysis, ICA)已經成功地運用在辨識的工作上。主值分析法是藉由對整組影像資料的全域共變異數矩陣(global covariance matrix)執行計算，所獲得的特徵向量是全域性的表現，它不適合運用在未調整之臉型上的辨識。相對地，獨立值分析法的特徵向量比主值分析的特徵向量較具有空間上的局部特性，而且可以得到較好的分類效能。在本論文中，主要的研究在評估主值分析法和獨立值分析法於臉型辨認上的抗雜訊能力。由於主值分析法和獨立值分析法辨識效能的比較和分析，是根據測試影像受到不同程度之高斯雜訊的干擾來評估；同時，我們也對特徵向量的選擇及相似度測量的方法進行評估的工作。從實驗結果分析，經過獨立值分析法與主值分析法之比較，我們可以得知，獨立值分析法具有較佳的抗雜訊能力，適合應用於具雜訊干擾環境下之臉型辨識工作。

Principal Component Analysis (PCA) and Independent Component Analysis (ICA) have been successfully used for pattern recognition. PCA is computed from the global covariance matrix of the full set of image data, the obtained basis vectors are global representations that are not suitable for the recognition of non-aligned faces. Relatively, ICA basis vectors are more spatially local than PCA basis vectors and give better face representation. This paper addresses the evaluation results of noise immunity for PCA and ICA in face recognition. The recognition performance of PCA and ICA are compared and analyzed when the test images are contaminated by noises. Also, different methods for eigenvector selection and similarity measures are evaluated. ICA has achieved better recognition performance in noise immunity than PCA, as shown in the experimental results. Therefore, ICA is applicable for face recognition under noise environment.

1. 緒論 1
1.1 研究動機與貢獻 2
1.2 章節架構 3
2. 臉型辨認系統 4
2.1 影像的前處理 5
2.1.1 臉型的抽取 5
2.1.2 亮度的一致性 6
2.2 特徵向量的選取 8
2.3 相似度和距離的測量 10
2.4 小結 13
3. 各種常用的統計方法 14
3.1 主值分析法 14
3.2 線性判別分析法 17
3.3 主值分析法與線性判別分析法結合的方法 19
3.4 獨立值分析法 21
3.4.1 統計獨立的基底影像 22
3.4.2 一個因數式編排的規則(a factorial code) 25
3.5 小結 28
4. 實驗 30
4.1 臉型資料庫 30
4.1.1 ORL臉型資料庫 30
4.1.2 FERET臉型資料庫 31
4.2 實驗一：亮度對臉型辨認的評估 33
4.3 實驗二：四種相似度測量方法的評估 36
4.4 實驗三：不同數量之特徵向量的評估 40
4.4.1 ORL臉型資料庫 40
4.4.2 FERET臉型資料庫 43
4.5 實驗四：抗雜訊力的分析（針對訓練樣本） 47
4.6 實驗五：抗雜訊力的分析（包含測試樣本） 52
4.6.1 ORL臉型資料庫 52
4.6.2 FERET臉型資料庫 55
4.7 小結 59
5. 結論 60
參考文獻 61
自 傳 64

[1] Rafael C. Gonzalez, and Richard E. Woods, Digital Image Processing, 2nd edition, 2002.
[2] Fukunaga K., Introduction to Statistical Pattern Recognition, Academic Press, 2nd edition, 1990.
[3] Chellappa R., Wilson C. L., and Sirohey S., “Human and Machine Recognition of Faces: a Survey,” Proceedings of the IEEE, Vol. 5, No. 83, pp. 705-740, 1995.
[4] Kirby M. and Sirovich L., “Application of the Karhunen-Loeve Procedure for the Characterization of Human Faces,” IEEE Trans. PAMI, Vol. 12, No. 1, pp. 103-108, 1990.
[5] M. Turk and A. Pentland., “Eigenfaces for Recognition,” Journal of Cognitive Neuroscience, Vol. 13, No. 1, pp. 71—86, 1991.
[6] Murase, H. and Nayar, S. K, “Illumination Planning for Object Recognition Using Parametric Eigen Space,” IEEE Trans. Pattern Anal. Machine Intell, Vol. 16, No. 12, pp. 1219-1227, 1994.
[7] Fisher, R. A., “The use of Multiple Measures in Taxonomic Problems,” Ann. Eugenics, No. 7, pp. 179-188, 1936.
[8] Swets D. and Weng J., “Using Discriminant Eigenfeatures for Image Retrieval,” IEEE Trans. PAMI, Vol. 18, No. 8, pp. 831-836, 1996.
[9] Swets D. and Weng J., “Hierarchical Discriminant Analysis for Image Retrieval,” IEEE Trans. PAMI, Vol. 21, No. 5, pp. 386-401, 1999.
[10] Belhumeur P., Hespanha J., and Kriegman D., “Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection,” IEEE Trans. PAMI, Vol. 19, No. 7, pp. 711-720, 1997.
[11] Aleix M. Martinez, and Avinash C. Kak., “PCA versus LDA,” IEEE Transactions on Pattern analysis and machine intelligence, Vol. 23, No. 2, February, 2001.
[12] Comon P., “Independent Component Analysis; A New Concept?” Signal Processing, Vol. 36, No. 3, pp. 287-314, 1994.
[13] Bell, A.J. and T.J. Sejnowski, “An Information-Maximization Approach to Blind Separation and Blind Deconvolution,” Neural Computation, Vol. 7, No. 6, pp. 1129-1159, 1995.
[14] Bell, A.J. and T.J. Sejnowski, “The 'Independent Components' of Natural Scenes are Edge Filters,” Vision Research, Vol. 37, No. 23, pp. 3327-3338, 1997.
[15] Bartlett, M. S. and T. J. Sejnowski, “Viewpoint invariant face recognition using independent component analysis and attractor networks,” in Neural Information Processing Systems-Natural and Synthetic, M. Mozer, M. Jordan, and T. Petsche, MIT Press: Cambridge, MA. pp. 817-823, 1997.
[16] Ming-Hsuan Yang, David Kriegman, and Narendra Ahuja, “Detecting Faces in Images: A Survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), Vol. 24, No. 1, pp. 34-58, 2002.
[17] Zhao W., Krishnaswamy A., Chellappa R., Swets D., and Weng J., “Discriminant Analysis of Principle Components for Face Recognition,” 3rd International Conference on Automatic Face and Gesture Recognition, pp. 336-341, 1998.
[18] Bartlett M. S., Lades H. M., and Sejnowski T. J., “Independent component representations for face recognition,” in SPIE Symposium on Electronic Imaging: Science and Technology; Conference on Human Vision and Electronic Imaging III. San Jose, CA., 1998.
[19] Liu, C. and H. Wechsler. “Comparative Assessment of Independent Component Analysis (ICA) for Face Recognition,” in International Conference on Audio and Video Based Biometric Person Authentication. Washington, D.C., 1999.
[20] Yambor W. S., “Analysis of PCA-Based and Fisher Discriminant-Based image recognition algorithms,” Technical report CS-00-013, Computer Science Department, Colorado State University, Fort Collins, July, 2000.