One Time Biometric Transform to secure biometric templates

Scalable Securing biometric information has become essential with growing biometric applications in different sectors of society. Vulnerability assessment plays a key role in improving the security of any security system by identifying the potential vulnerabilities and proposing countermeasures to mitigate the threats posed by them. In this work self-generated and dynamic helper data based system is proposed to encrypt the biometric templates. We name this methodology as `One Time Biometric Transform´ (OTBT). Biometric information is statistically learned and probabilistic matching is performed to discriminate genuine from imposters. The system was tested using CASIA iris database and by probabilistic matching an EER of 1.96% is achieved. Strength analysis of the system for three different challenging databases is also presented.