臺灣博碩士論文加值系統

人臉辨識於現今生活中扮演相當重要的角色，在許多領域中都可被應用。至今有許多關於人臉辨識的研究誕生，大部分都在處理單一影響辨識的因素，例如亮度、姿勢或表情的變化、影像遮蔽等因素。
本篇論文主要以同時處理亮度變化以及影像遮蔽為目標，運用著名基於sparse representation分類法(SRC)結合改良過的weber 區域梯度描述子(VWLGD)。透過VWLGD的處理能有效地降低亮度變化的影響，再將處理後的影像透過SRC進行分類以達到實驗目的。由實驗結果可以看出本篇論文所提出的方法有良好的辨識率，對於亮度變化和影像遮蔽也都能有效的處理。

Face recognition plays an important role nowadays. It is critical in a wide range of applications such as mug-shot database matching, credit card verification, security system, and scene surveillance. A large amount of works have been done in face recognition. Most of them deal with uncontrolled variations such as changes in illumination, pose, expression and occlusion individually.
　　In this thesis, variable illumination and occlusion are mainly discussed. We propose an approach combining sparse representation based classification (SRC) and varied weber local gradient descriptor (VWLGD) to deal with them. It is effective in variable illumination by using VWLGD. Then the processed images can be classified with SRC to achieve the goals. Experimental results show that this method achieves high recognition rates, and is quite effective in variable illumination and occlusion.

Contents
Chapter 1 Introduction 1
1.1 Methodology 1
1.2 Retinal Modeling 2
1.2.1 Light adaption filter 3
1.2.2 OPL filter 4
1.3 Weber Local Descriptor 5
1.3.1 Weber’s Law 5
1.3.2 Differential Excitation 6
1.3.3 Discussion on Using Arctangent function 8
1.4 Classifier 10
1.5 Thesis Organization 13
Chapter 2 Related Work 14
2.1 Improved Retinal Modeling (IRM) 14
2.1.1 Retinal Illumination Nonlinear Adaption 14
2.1.2 Local Illumination Estimation 15
2.1.3 Illumination Classification 15
2.1.4 Proposed Improve Retinal Modeling 16
2.2 Weber Local Gradient Descriptor (WLGD) 18
2.2.1 Proposed WLGD 18
2.2.2 Properties of WLGD 19
2.3 Sparse Representation based Classifier 21
2.3.1 Test Sample as a Sparse Linear Combination of Training Samples 21
2.3.2 Dealing with Occlusion 22
2.3.3 Classification Based on Sparse Representation 23
Chapter 3 Proposed Method 25
3.1 Framework 25
3.2 Varied Weber Local Gradient Descriptor (VWLGD) 26
3.3 Combining SRC with VWLGD 29
3.4 VWLGD Division 30
Chapter 4 Experimental Results and Discussions 32
4.1 Face Database 32
4.2 Experimental Results 33
4.2.1 Facial Expression 33
4.2.2 Illumination Change 34
4.2.3 Occlusion 35
4.2.4 Illumination + Occlusion 36
4.2.5 Using Two Training Samples to Recognition 37
4.2.6 Different Division Levels 38
4.2.7 Results on Extended Yale B face database 39
Chapter 5 Conclusions 42
Reference 43

[1]Ngoc-Son Vu and Alice Caplier, “Illumination-Robust Face Recognition Using Retina Modeling, 16th IEEE International Conference on Image Processing (ICIP), 2009, pp. 3289-3292.
[2]A. Benoit, “The Human Visual System as a Complete Solution for Image Processing, Ph.D. thesis, INPG, Grenoble, France, 2007.
[3]Jie Chen, Shiguang Shan, Chu He, Guoying Zhao, Matti Pietikainen, Xilin Chen and Wen Gao, “WLD: A Robust Local Image Descriptor, IEEE Transactions on Pattern Analysis and Machine Intelligence (Volume:32 , Issue: 9 ) ,IEEE Biometrics Compendium, IEEE RFIC Virtual Journal, IEEE RFID Virtual Journal, 2009, pp. 1705-1720
[4]A.K. Jain, “Fundamentals of Digital Image Processing, PrenticeHall, 1989, pp. 51.
[5]John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry and Yi Ma, “Robust Face Recognition via Sparse Representation, IEEE Transactions on Pattern Analysis and Machine Intelligence (Volume:31 , Issue: 2 ), 2008, pp. 210-227.
[6]R. Duda, P. Hart and D. Stork, “Pattern Classification, second ed. John Wiley ＆ Sons, 2001.
[7]J. Ho, M. Yang, J. Lim, K. Lee and D. Kriegman, “Clustering Appearances of Objects under Varying Illumination Conditions, Proc. IEEE Int’l Conf. Computer Vision and Pattern Recognition, 2003, pp. 11-18.
[8]S. Li and J. Lu, “Face Recognition Using the Nearest Feature Line Method, IEEE Transactions on Neural Networks (Volume:10 , Issue: 2 ), 1999, pp. 439-443.
[9]Yong Cheng, Liangbao Jiao, Zuoyong Li and Xuehong Cao, “An Improved Retinal Modeling for Illumination Face Recognition, IEEE International Conference on Image Processing (ICIP), 2014, pp. 244-247.
[10]Shanshan Fang, Jucheng Yang, Na Liu and Yarui Chen, “Weber Local Gradient Pattern (WLGP) Method for Face Recognition, Lecture Notes in Computer Science, Vol. 9428, 2015, pp. 186-192.
[11]Biao Wang, Weifeng Li, Wenming Yang and Qingmin Liao , “Illumination Normalization Based on Weber’s Law with Application to Face Recognition, IEEE Signal Processing Letters (Volume:18 , Issue: 8 ) , 2011, pp. 462-465.
[12]S. Boyd and L. Vandenberghe, “Convex Optimization, Cambridge Univ. Press, 2004.
[13]E. Candes and J. Romberg, “l1-Magic: Recovery of Sparse Signals via Convex Programming, http://www.acm.caltech.edu//l1magic/, 2005.
[14]A.M. Martinez and R. Benavente, “The AR Face Database, Technical report, CVC Technical report, no. 24, 1998.
[15]A. Georghiades, P. Belhumeur and D. Kriegman, “From Few to Many: Lighting Cone Models for Face Recognition under Variable Lighting and Pose, IEEE Trans. Pattern Anal. Mach. Intell., Vol. 23, pp. 643-660, 2011.
[16]J. Anderson, “An Introduction to Neural Networks, The MIT Press, 1995.
[17]M. Turk and A. Pentland, “Eigenfaces for Recognition, Proc. IEEE Int’l Conf. Computer Vision and Pattern Recognition, 1991.
[18]P. Belhumeur, J. Hespanda and D. Kriegman, “Eigenfaces versus Fisherfaces: Recognition Using Class Specific Linear Projection, IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 19, no. 7, pp. 711-720, July 1997.
[19]X. He, S. Yan, Y. Hu, P. Niyogi and H. Zhang, “Face Recognition Using Laplacianfaces, IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 27, no. 3, pp. 328-340, Mar. 2005.