Authentication systems usually adopt either theconventional identifier-password paradigm or different kinds oftokens (e.g., badges, keys). However, passwords can be disclosedwhile being input and tokens can be stolen and used by impostors.As a result, in the last decades biometric techniques were developedto identify a user through physiological features that cannot be stolenor counterfeited. However, even those techniques have their flaws,and for this reason recent research addressed the combination ofmultiple identification factors. In this context, this work proposesVisilabFaceRec, a multi factor authentication system based on thecombination of a dual-stage cascading classifier for biometricidentification (face recognition) with an encrypted RFID tag for token-based authentication. Unlike other approaches in the literature thatpropose a centralized database for storing biometric data, withserious risks regarding user privacy, our work avoids a centralizeddatabase and stores sensitive data in the RFID, thus also making thesystem performance independent of the total number of subjectsenrolled. The proposed architecture is able to simultaneouslyminimize the False Acceptance Rate and the False Rejection Rate,thanks to an innovative approach for the calculation of the decisionthresholds for the two discriminators. VisilabFaceRec has beenrealized on a commercial board for embedded computing andproven to be able to run in near real-time. The paper describes thesystem architecture and the algorithm used to jointly determine thecouple of decision thresholds for the cascading classifiers, andproposes a performance evaluation, in terms of both accuracy andspeed, on a well-known and publicly available face database.

A Biometric Authentication System Based on Face Recognition and RFID tags

LO BELLO, Lucia
2014-01-01

Abstract

Authentication systems usually adopt either theconventional identifier-password paradigm or different kinds oftokens (e.g., badges, keys). However, passwords can be disclosedwhile being input and tokens can be stolen and used by impostors.As a result, in the last decades biometric techniques were developedto identify a user through physiological features that cannot be stolenor counterfeited. However, even those techniques have their flaws,and for this reason recent research addressed the combination ofmultiple identification factors. In this context, this work proposesVisilabFaceRec, a multi factor authentication system based on thecombination of a dual-stage cascading classifier for biometricidentification (face recognition) with an encrypted RFID tag for token-based authentication. Unlike other approaches in the literature thatpropose a centralized database for storing biometric data, withserious risks regarding user privacy, our work avoids a centralizeddatabase and stores sensitive data in the RFID, thus also making thesystem performance independent of the total number of subjectsenrolled. The proposed architecture is able to simultaneouslyminimize the False Acceptance Rate and the False Rejection Rate,thanks to an innovative approach for the calculation of the decisionthresholds for the two discriminators. VisilabFaceRec has beenrealized on a commercial board for embedded computing andproven to be able to run in near real-time. The paper describes thesystem architecture and the algorithm used to jointly determine thecouple of decision thresholds for the cascading classifiers, andproposes a performance evaluation, in terms of both accuracy andspeed, on a well-known and publicly available face database.
2014
Authentication; Biometrics; RFID tags; Security
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11769/34659
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 19
  • ???jsp.display-item.citation.isi??? ND
social impact