Construction and Performance of Robust Fingerprint Key Extractor

摘要

Biometric cryptosystem can not only provide an efficient mechanism for template protection, but also facilitate cryptographic key management, thus becomes a promising direction in information security field. Though a number of bio-crypto algorithms have been proposed, they have limited practical applicability due to poor performance and weak security of the biometric template. Since Yevgeniy Dodis first proposed the concept of fuzzy extractor, many theoretical work have been done on this project, but no concrete scheme have been proposed which can be used in a real cryptographic application. In this paper, we first give the concept of Robust Fingerprint Key Extractor (RFKE), then fingerprint's FingerCode model, which capture ridge orientation and frequency information in a fingerprint, and fingerprint's minutiae model, which are fingerprint's local ridge characteristics, are incorporated together to construct a practical RFKE. This is the first non-trivial Fingerprint Key Extraction (FKE) scheme which can stand against active adversary. The proposed scheme is robust towards stored biometric template attacks in common reference string (CRS) model. Experiments on two public fingerprint databases, UPEK database of Institute of Automation of the Chinese Academy of Sciences (CASIA) and DB2 database of FVC2000, show that the use of vault reduces the Impostor Success Rate (ISR) of the system significantly when coding dimension is low, while the Genuine Success Rate (GSR) drops only a little no matter coding dimension is low or high.