There are tremendous need increase for personal verification and identification in internet security, electronic commerce and access control in recent years. Also, as the demands for security in many applications such as data protection and financial transaction become an increasingly relevant issues, the importance of biometric technology is rapidly increasing. In this paper, we explored face recognition system for person authentication. We explicitly state the design decisions by introducing a generic modular PCA face recognition system. We designed implementations of each module, and evaluate the performance variations based on virtual galleries and probe sets. We perform experiments and report results using equal error rates (EER) based on verification scenario for access control applications.
展开▼
