Abstract
Due to the performance limitation of using a single biometric, many researchers have been investigating ways to integrate multiple biometrics for the purpose of
authentication. To this end, there have been numerous articles on promoting multiple biometric identifiers. It is thus important to know when integrating improves the performance and when it does not. This research focuses on integrating biometrics at the decision level. In this article, the formulas are first derived for the probabilities of a False Accept and a False Reject when n Biometric tests are combined. The conditions are then obtained for performance improvement and degradation of combing two biometric identifiers. To the end, some other interesting results of integrating biometric solutions are also presented.
authentication. To this end, there have been numerous articles on promoting multiple biometric identifiers. It is thus important to know when integrating improves the performance and when it does not. This research focuses on integrating biometrics at the decision level. In this article, the formulas are first derived for the probabilities of a False Accept and a False Reject when n Biometric tests are combined. The conditions are then obtained for performance improvement and degradation of combing two biometric identifiers. To the end, some other interesting results of integrating biometric solutions are also presented.
Original language | American English |
---|---|
Pages | 1-6 |
Number of pages | 6 |
State | Published - May 2006 |
Externally published | Yes |
Event | Florida Conference on Recent Advances in Robotics (FCRAR) - Miami, United States Duration: May 25 2006 → May 26 2006 |
Conference
Conference | Florida Conference on Recent Advances in Robotics (FCRAR) |
---|---|
Abbreviated title | FCRAR |
Country/Territory | United States |
City | Miami |
Period | 5/25/06 → 5/26/06 |