Motion Compensation for Face Recognition Based on Active Differential Imaging

Active differential imaging has been proved to be an effective approach to remove ambient illumination for face recognition. In this paper we address the problem caused by motion for a face recognition system based on active differential imaging. A moving face will appear at two different locations in the ambient illumination frame and combined illumination frame and as result artifacts are introduced to the difference face image. An approach based on motion compensation is proposed to deal with this problem. Experiments on moving faces demonstrate that the proposed approach leads to significant improvements in face identification and verification results.

[1]  Bernadette Dorizzi,et al.  Active differential CMOS imaging device for human face recognition , 2006, IEEE Signal Processing Letters.

[2]  Josef Kittler,et al.  Face Recognition Using Active Near-IR Illumination , 2005, BMVC.

[3]  Michael J. Black,et al.  The Robust Estimation of Multiple Motions: Parametric and Piecewise-Smooth Flow Fields , 1996, Comput. Vis. Image Underst..

[4]  Jiri Matas,et al.  On Combining Classifiers , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Azriel Rosenfeld,et al.  Face recognition: A literature survey , 2003, CSUR.

[6]  Y. Lida,et al.  A 100 frame/s CMOS active pixel sensor for 3D-gesture recognition system , 1999, 1999 IEEE International Solid-State Circuits Conference. Digest of Technical Papers. ISSCC. First Edition (Cat. No.99CH36278).

[7]  Shimon Ullman,et al.  Face Recognition: The Problem of Compensating for Changes in Illumination Direction , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Bernadette Dorizzi,et al.  Specific Sensors for Face Recognition , 2006, ICB.

[9]  Yang Ni,et al.  CMOS active differential imaging device with single in–pixel analog memory , 2002 .

[10]  Josef Kittler,et al.  Ambient Illumination Variation Removal by Active Near-IR Imaging , 2006, ICB.