Biometric Mobile Template Protection: A Composite Feature Based Fingerprint Fuzzy Vault

Biometric authentication is emerging as the promising solution to conventional cryptography based authentication technologies. However, protecting users´ biometric templates stored in a mobile device in a secure way is a challenge issue and has attracted many attentions. As one of the possible solutions, the Fuzzy Vault construct binds a secret key and biometric information to provide template protection. Most existing Fingerprint Fuzzy Vault algorithms use pre-aligned fingerprint impressions and rely strongly on image registration, a process that is well known to be nontrivial and unreliable. Moreover, it is inherently insecure to store raw fingerprint images for the alignment. In this paper, we propose a Fingerprint Fuzzy Vault based on composite features which are reliable, distortion tolerant and registration-free. Experimental results on public database show that our scheme can improve verification performance significantly.