本論文首次應用光源正規化(Illumination Normalization)及局部特徵(Component Features)改進稀疏表示法(Sparse Representation) 之人臉辨識，並稱所提出的方法為INCO(Illumination-Normalized and Component-Oriented)稀疏表示法。不同於目前的稀疏表示法藉由多樣性的訓練樣本建立所需的線性基底，INCO 稀疏表示法先將訓練樣本進行光源正規化，再透過局部特徵的擷取，增強線性基底對於變化光源的穩健度。為結合局部(Component)與全域(Holistic)特徵，本研究提出了兩層式(Bilayer)和晶格式切割(Dense-on-Features)的結合方式並評估其效益，也比較了多種光源正規化的機制，不同的局部區塊，不同的影像特徵與特徵抽取之相關參數，結論出INCO稀疏表法之最佳系統架構與參數設置。透過不同標準資料庫(FERET、FRGC、Multi-PIE)與目前較佳之人臉辨識方法進行效能評比，本研究證實所提出之方法具有高度競爭力。

We merge illumination normalization and component features into the framework of
Sparse Representation-based Classification (SRC) for face recognition across illumination, and name it Illumination-Normalized and Component-Oriented (INCO) SRC. Unlike most SRC-based face recognition which constructs a dictionary from a training set with sufficient illumination variation, the proposed method first adopts a dictionary with illumination normalized training set and then extract the component features. For further improving the performance, we propose two combination schemes, Bilayer and Dense-on-Features, to combine component feature with holistic feature. We also compare different illumination normalizations, local region, and feature extraction to obtain the optimal setting to construct the framework of INCO-SRC. Experiments on FERET, FRGC and Multi-PIE databases show that the performance of the proposed method can be competitive to the state of the art.

[1] A. J. K. Bonnen, B.F. Klare, "Component-based representation in automated face recognition," Information Forensics and Security, IEEE Trans. on, vol. 8, p. 239–253, Jan. 2013.
[2] J. G. W. Deng, J. Hu, "Extended src : Undersampled face recognition via intraclass variant dictionary," PAMI, IEEE Trans. on, vol. 34, p. 1864–1870, Sept. 2012.
[3] A. Wagner, J. Wright, A. Ganesh, H. Mobahi, Z. Zhou, and Y. Ma, "Towards a
practical face recognition system: Robust registration and illumination by sparse representation," PAMI, IEEE Trans. on, vol. 34, p. 372–386, Feb. 2012.
[4] X. Tan and B. Triggs, "Enhanced local texture feature sets for face recognition under difficult lighting conditions," Image Processing, IEEE Trans. on, vol. 19, p. 1635–1650, June 2010.
[5] J. H. P. Belhumeur and D. Kriegman, "Eigenfaces vs fisherfaces:recognition using class specific linear projection," PAMI, IEEE Trans. on, vol. 19, p. 711–720, July 1997.
[6] W. Y. B. Wang, W. Li and Q. Liao, "Illumination normalization based on weber's law with application to face recognition," Signal Processing Letters, vol. 18, p. 462–465, Aug. 2011.
[7] J. K. Y.K. Park, S.L. Park, "Retinex method based on adaptive smoothing for illumination invariant face recognition," Signal Processing, vol. 88, p. 1929–1945, 2008.
[8] P. Phillips, P. Flynn, T. Scruggs, K. Bowyer, J. Chang, K. Hoffman, J. Marques, J. Min, and W. Worek, "Overview of the face recognition grand challenge," CVPR, vol. 1, p. 947–954, June 2005.
[9] Z. R. D. Jobson and G. Woodell, "A multiscale retinex for bridging the gap between color images and the human observation of scenes," Image Processing, IEEE Trans. on, vol. 6, p. 965–976, July 1997.
[10] T. Sim, S. Baker, and M. Bsat, "The cmu pose, illumination, and expression
database," PAMI, IEEE Trans. on, vol. 25, p. 1615–1618, Dec. 2003.
[11] "INFace Toolbox." http://luks.fe.uni-lj.si/sl/osebje/vitomir/face_
tools/INFace/refs.html.
[12] J. Zou, Q. Ji, and G. Nagy, "A comparative study of local matching approach for face recognition about," Image Processing, IEEE Trans. on, vol. 16, p. 2617–2628, Oct. 2007.
[13] A. G. J. Wright, A.Y. Yang, S. Sastry, and Y. Ma, "Robust face recognition via sparse representation," PAMI, IEEE Trans. on, vol. 31, p. 210–227, Feb. 2009.
[14] P. Phillips, H. Moon, P. Rizvi, and P. Rauss, "The feret evaluation method for face recognition algorithms," PAMI, IEEE Trans. on, vol. 22, p. 1090–1104, Oct. 2000.
[15] Y. O. P. Sinha, B. Balas and R. Russell, "Face recognition by humans: Nineteen results all computer vision researchers should know about," Proceedings, vol. 94, p. 1948–1962, Nov. 2006.
[16] X. Yong, D. Zhang, J. Yang, and J.-Y. Yang, "A two-phase test sample sparse representation method for use with face recognition," Circuits and Systems for Video Technology, IEEE Trans. on, vol. 21, p. 1255–1262, Sept. 2011.
[17] A. Y. Yang, S. S. Sastry, A. Ganesh, and Y. Ma, "Fast l1-minimization algorithms and an application in robust face recognition: A review," in ICIP, p. 1849–1852, 2010.
[18] N. Vu and A. Caplier, "Enhanced patterns of oriented edge magnitudes for face recognition and image matching," Image Processing, IEEE Trans. on, vol. 21, p. 1352–1365, Mar. 2012.
[19] J. Yang, K. Yu, and T. Huang, "Supervised translation-invariant sparse coding," CVPR, p. 3517–3524, June 2010.
[20] C. H. Chan, J. K. M. A. Tahir, and M. Pietikainen, "Multiscale local phase quantization for robust component-based face recognition using kernel fusion of multiple descriptors," PAMI, IEEE Trans. on, vol. 35, p. 1164–1177, May 2013.
[21] Y. Su, S. Shan, X. Chen, and W. Gao, "Hierarchical ensemble of global and local classifiers for face recognition," Image Processing, IEEE Trans. on, vol. 18, no. 8, p. 1885–1896, 2009.
[22] P. Lee, S. Wu, and Y.-P. Hung, "Illumination compensation using oriented local histogram equalization and its application to face recognition," Image Processing, IEEE Trans. on, vol. 21, no. 9, p. 4280–4289, 2012.
[23] T. Cootes, C. Tayler, and D. Cooper, "Active shape models—their training and application," Comput. Vis. Image Understand., vol. 61, p. 38–59, 1995.