Decision Level Fusion of Fingerprint Minutiae Based Pseudonymous Identifiers

In a biometric template protected authentication system, a pseudonymous identifier is the part of a protected biometric template that can be compared directly against other pseudonymous identifiers. Each compared pair of pseudonymous identifiers results in a verification decision testing whether both attributes are derived from the same individual. Compared to an unprotected system, most existing biometric template protection methods cause to a certain extent, degradation in biometric performance. Therefore fusion is a promising method to enhance the biometric performance in template protected systems. Compared to feature level fusion and score level fusion, decision level fusion exhibits not only the least fusion complexity, but also the maximum interoperability across different biometric features, systems based on scores, and even individual algorithms. However, performance improvement via decision level fusion is not obvious. It is influenced by both the dependency and the performance gap among the conducted tests for fusion. We investigate in this paper several scenarios (multi-sample, multi-instance, multi- sensor, and multi-algorithm) when fusion is performed on binary decisions obtained from verification of fingerprint minutiae based pseudonymous identifiers. We demonstrate the influence on biometric performance from decision level fusion in different fusion scenarios on a multi-sensor fingerprint database.