Un sistema di autenticazione biometrica basato su tag RFID e riconoscimento del viso

This work proposes a Multi Factor Authentication system based on a dual-stage cascading classifier for biometrics identification (face recognition) and an RFID tag. A novel approach has been devised in order to automatically determine the thresholds for the two stages. The system records the biometric data of each user directly on his/her RFID tag and offers several benefits. First, as the biometrics data are stored in a database embedded in the RFID data, there is no need for a centralized database or server storing biometrics data and this solves a number of legal and technological issues relevant to data privacy. Second, the size of the embedded database used for authentication is significantly reduced, to the benefit of both speed and accuracy of the identification process. Finally, the proposed system is able to operate at very low resolution (down to 20x15 pixel) in near-real time and therefore it is suitable for application to restricted areas access control (being relatively cheap, it may be applied even to house access control), authentication for confidential services such as electronic payments, and homeland security. The system has been realized on a board for embedded systems. This paper describes the authentication system and discusses its performance in terms of accuracy and authentication time.