Fuzzy Comparison Scheme in Biometric Template Matching

Abstract

The inherent variation of the same biometric makes it difficult to replace passwords directly with biometrics in a cryptographic scheme. Comparison of the true sets is a key step in biometric template matching or encryption. In FCS (short for fuzzy comparison scheme), an algorithm for biometric template matching is proposed based on fuzzy theory. The proposed method defines two fuzzy sets, measures the nearness between the two true sets, and selects an available threshold. Moreover, this method can be generalized into high dimension space and multi-set matching. Algorithm implementation and numerical example show that the fuzzy comparison scheme is effective and tolerant because of the threshold, which can make it more explicit for uncertainty.