Cryptographic key generation from multimodal template using fuzzy extractor

Abstract

The encryption techniques, biometrics and cryptography integrate to form biometric cryptosystems. These are designed either to bind a cryptographic key or to generate cryptographic key using biometric features. The deployment of bio-cryptosystem technique, namely fuzzy extractor in multimodal biometric system leads to increase in user privacy and system security. This paper provides with a framework where feature level fusion of iris and dual fingerprint forms a multimodal template and key is generated using fuzzy extractors, in order to provide reliability and good recognition performance. The hash function is used to protect the key generated from biometric traits. Since fuzzy extractor operates only on ordered dataset. However the minutiae points of fingerprint are unordered, so an algorithm is designed for conversion of unordered minutiae points to ordered minutiae dataset to make it consistent for key generation methods.