Secure Palmprint Verification System
Mauricio Ramalho · 2010
The usage of biometric characteristics (e.g., palmprints, fingerprints, irises) for individual recognition is becoming increasingly popular and many applications are already available. This form of recognition has many advantages over the traditionally used methods (e.g., password, smart card), because a biometric characteristic cannot be forgotten or shared, as it is intrinsically associated with the individual. However, due to this property, some security issues arise. One of the most concerning vulnerabilities in biometric recognition systems is that biometric templates, when stored in an unencrypted form, are an easy target for an attack. Once a biometric template is jeopardized, it cannot be reissued and the biometric characteristic associated to that template is compromised forever. An attacker could then gain access to all the services, applications or areas that use the same biometric characteristic. Since each individual has a limited number of biometric characteristics, some measures must be taken to overcome the security issues in biometric recognition systems. This dissertation presents a biometric recognition system with a secure template storage scheme. The generation of different templates based on the same biometric data is supported, as well as cancellable templates. Furthermore, it is not possible to recover the original biometric data from the data stored in the system’s database.