The face recognition accuracy degrades under varying illumination and pose.
Many researches have discussed solutions to solve the two problems, but most of them need multiple training images. Unfortunately, this requirement cannot always be satisfied. In this thesis, we have proposed a system for face recognition called automatic pose normalization and edge map recognizer (APER). The APER employs pose normalization and edge map for face recognition with single training image. Pose normalization by morphing before PCA recognition is proposed to overcome the aspect angle problem effectively and an automatic corresponding algorithm has been derived to enhance the efficiency of the pose normalization process. Edge map with level-mask process used in face recognition is less sensitive in illumination variation condition and more is more efficient in computation. In conclusion, the APER can improve the performance of conventional PCA approach under varying pose and illumination with single training image. The comparison of the APER and the related works is shown in Table 7.1.
Table 7.1 The comparison of related works and APER
To improve the performance and the robustness of the APER, some enhancements can be done in the future:
(a) We are currently extending the system based on other recognition approaches. We believe that pose normalization and edge map are also useful for other face recognition approaches, not only for PCA.
(b) Edge map can also be extended to derive the automatic correspondence algorithm for pose normalization, and the edge map can be morphed for face recognition directly. Morphing by the edge map will make the system more efficient and we will prove it in the future.
(c) The automatic correspondence algorithm sometimes extracts error correspondence, and the error correspondence degrades the performance of face recognition. Therefore, the recognition rate of APER is lower than PER. We will improve our automatic correspondence algorithm as accurate as hand-marked.
(d) In this thesis, we focus on intensity images but not color images. Pose normalization, edge map and automatic correspondence can directly be adopted on color images, but PCA recognition method have to be modified.
Such modification for color images will be implemented in the future.
Reference
[1] W. Zhao, R. Chellappa, and A. Rosenfeld, P.J. Phillips, “Face Recognition: A Literature Survey,” ACM Computing Surveys, 399-458, 2003
[2] P. J. Phillips, R. M. McCabe, and R. Chellappa. “Biometric image processing and recognition.” Proc. of the European Signal Processing Conference, 1998.
[3] Matthew A. Turk and Alex P. Pentland, “Face Recognition Using Eigenfaces,”
IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 586-591, 1991
[4] Rajkiran Gottumukkal , and Vijayan K. Asari, “An improved face recognition technique based on modular PCA approach,” Pattern Recognition Letters, Vol.25 No.4, 429-436,2004
[5] Y. Adini, Y. Moses, and S. Ullman, “Face recognition: The problem of compensating for changes in illumination direction.” IEEE Transaction on Pattern Analysis and Machine Intelligence, Vol.19, 721-732, 1997
[6] Y.li, S. Gong, and H. Liddell, “Support vector regression and classification based multi-view face detection and recognition,” Proc. of Automatic Face and Gesture Recognition Conference, Grenoble, 300-305, 2000.
[7] T. Riklin-Raviv and A. Shashua, “The quotient image: Class based rerendering and recognition with varying illuminations.” Proc. of IEEE Conference on Computer Vision and Pattern Recognition, 566-571, 1999
[8] Georghiades, A. S., Kriegman, D. J., and Belhumeur, P. N., “Illumination cones for recognition under variable lighting: Faces” , Proc. of IEEE Conference on Computer Vision and Pattern Recognition, 52-58, 1998
[9] Fukunaga, Keinosuke, “Introduction to Statistical Pattern Recognition,”
Elsevier,1990
[10] B. Manjunath, R. Chellappa, and C. Malsburg, "A feature based approach to face recognition," Proc. of IEEE Conference on Computer Vision and Pattern Recognition, 373-378, 1992
[11] Cox, J. Ghosn, and P. Yianilos, "Feature-based face recognition using mixture-distance," Proc. of IEEE Conference on Computer Vision and Pattern Recognition, 209-216, 1996
[12] F. Shamaria and S. Young, "HMM based architecture for face identification,"
Image and Vision Computing, Vol.12, 537-583, 1994
[13] A. Nefian and M. Hayes, "Hidden Markov models for face recognition," Proc. of International Conference on Acoustics, Speech and Signal Processing, 2721-2724, 1998
[14] B. Moghaddam and A. Pentland, "Probabilistic visual learning for object representation," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.19, 696-710,1997
[15] D. Swets and J. Weng, "Using discriminant eigenfeatures for image retrieval,"
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 18, 831-836, 1996
[16] P. N. Belhumeur, J. Hespanha, and D. J. Kriegman, "Eigenfaces vs. Fisherfaces:
Recognition using class specific linear projection," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.19, 711-720, 1997
[17] W. Zhao, R. Chellappa, and A. Krishnaswamy, "Discriminant analysis of principal components for face recognition," Proc. of International Conference on Automatic Face and Gesture Recognition, 336-341, 1998
[18] P.Phillips, P.Rauss, and S.Der.Feret (face recognition technology) recognition algorithm development and test report. Technicaul Report ARL 995, U.S. Army Research Laboratary, 1996
[19] P.J.Phillips, P. Grother, R. Micheals, D. Blackburn, E. Tabassi, and J. Bone.
Face recognition vendor test 2002: Evaluation report. NISTIR 6965, National Institute of Standards and Technology http://www.frvt.org.
[20] A. Pentland, B. Moghaddam, and T. Starner, "View-based and modular eigenspaces for face recognition," Proc. of IEEE Conference on Computer Vision and Pattern Recognition, 84-91,1994
[21] T. Cootes, K. Walker, and C. Taylor, "View-based active appearance models,"
Proc. of International Conference on Automatic Face and Gesture Recognition, 227-238, 2000
[22] K. Sung and T. Poggio, "Example-based learning for view-based human face detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.19, 775-779, 1997
[23] A. Georghiades, and P. BelhumeurFrom, "Few to many：Illumination cone models for face recognition under variable lighting and pose," IEEE Transactions on Pattern Analysis and Machine Intelligence,Vol.23, No. 6, 643～660, 2001 [24] Acharya and Ray, “Image Processing: Principles and Applications,”
Wiley-Interscience, 2005
[25] Thaddeus Beier and Shawn Neely, “Feature-Based Image Metamorphosis,”
Pro. of Annual Conference on Computer Graphics and Interactive Techniques (SIGGRAPH). Vol. 26, 35-42, 1992 .
[26] Alam, M.S.; Al-Samman, A.R.; “Ultrafast invariant face recognition.” Proc. of IEEE. National Aerospace and Electronics Conference, 486-493 ,2000
[27] James Matthews. “An introduction to edge detection: The Sobel edge detector.”
Available at http://www.generation5.org /content/2002/im01.asp, 2002.
[28] Angus M. K. Siu and Rynson W. H. Lau, “Image Registration for Image-Based Rendering”, IEEE Transactions on Image Processing, Vol. 14, No. 2, 241-252, 2005
[29] Winston Li and Henry Leung, “A Maximum Likelihood Approach for Image Registration Using Control Point And Intensity”, IEEE Transactions on Image Processing, Vol. 13, No. 8, 1115- 1127, 2004
[30] Huang, J, and Wechsler, H.; “Visual routines for eye location using learning and evolution”, IEEE Transactions on Evolutionary Computation, Vol. 4, No. 1, 73 – 82, 2000
[31] Paperno, E, and Semyonov, D; ”A new method for eye location tracking”, IEEE Transactions on Biomedical Engineering, Vol 50, No. 10, 1174 – 1179, 2003 [32] Jung-Il Choi, Che-Woo La, Phill-Kyu Rhee, and Yong-Lai Bae; “Face and eye
location algorithms for visual user interface”, IEEE First Workshop on Multimedia Signal Processing, 239 – 244, 1997
[33] Wiskott, L, Fellous, J.-M, Kuiger, N, and Malsburg, C “Face recognition by elastic bunch graph matching”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19, No. 7, 775 – 779, 1997
[34] Mokhtarian, F, and Suomela, R. ”Curvature scale space for robust image corner detection”, Proc. of International Conference on Pattern Recognition, Vol. 2, 1819-1821, 1998
[35] P. Saeedi, D. Lowe, and P.Lawrence, “An efficient binary corner detector,”
Proc. of International Conference on Control, Automation, Robotics And Vision, Vol. 1, 338- 343, 2002,
[36] S. Alkaabi, and F. Deravi , “Iterative Corner Extraction and Matching for Mosaic Construction”, Proc. of the Canadian Conference on Computer and Robot Vision, 468- 475, 2005
[37]Han Wang and Brady, M. “A practical solution to corner detection”, Proc. of IEEE Conference on Image Processing, Vol. 1, 919-923, 1994
[38] ORL face database, AT&T Laboratories, Cambridge, U.K. [Online].
Available at http://www.cam-orlco.uk/facedatabase.html