A Novel Template Protection Algorithm for Iris Recognition

As one of biometric technologies for personal identification and verification, many algorithm for iris recognition are proposed in recent decades. In these algorithms, iris templates are stored uncovered in the database. Once the database is compromised, the attackers can potentially obtain sufficient information to impersonate the users. This paper proposes a novel template protection algorithm for iris recognition, which stores the hash value instead of the iris template. Common vector method and non-uniform quantization method are used to eliminate the contradiction between the fuzzy character of iris information and the sensitive of the hash function. Experiments show the performance of the iris recognition system with template protection is improved.