*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Face Recognition for Vehicle Personalization
Committee:
Dr. David Anderson, ECE, Chair , Advisor
Dr. Monson Hayes, ECE, Co-Advisor
Dr. Biing-Hwang Juang, ECE
Dr. Aaron Lanterman, ECE
Dr. Edward Coyle, ECE
Dr. Mark Smith, Swedish Royal Institute of Technonogy
Abstract:
The objective of this dissertation is to develop a system of practical technologies to implement an illumination robust, consumer grade biometric system based on face recognition to be used in the automotive market. Most current face recognition systems are compromised in accuracy by ambient illumination changes. Especially outdoor applications including vehicle personalization pose the most challenging environment for face recognition. The point of this research is to investigate practical face recognition used for identity management in order to minimize algorithmic complexity while making the system robust to ambient illumination changes. We start this dissertation by proposing an end-to-end face recognition system using near infrared (NIR) spectrum. The advantage of NIR over visible light is that it is invisible to the human eyes while most CCD and CMOS imaging devices show reasonable response to NIR. Therefore, we can build an unobtrusive night-time vision system with active NIR illumination. In day time the active NIR illumination provides more controlled illumination condition. Next, we propose an end-to-end system with active NIR image differencing which takes the difference between successive image frames, one illuminated and one not illuminated, to make the system more robust on illumination changes. Furthermore, we addresses several aspects of the problem in active NIR image differencing which are motion artifact and noise in the difference frame, namely how to efficiently and more accurately align the illuminated frame and ambient frame, and how to combine information in the difference frame and the illuminated frame. Finally, we conclude the dissertation by citing the contributions of the research and discussing the avenues for future work.
