Regresar

Evaluation of Cancellable Biometric Schemes Applied to Iris Recognition in User Authentication and Identification Tasks

Abstract:

Biometric systems are a powerful alternative for user authentication and identification, since they are based on the usage of some physical or behavioral trait. However, biometric traits are generally immutable, resulting in a permanent compromise when a user template is shared or stolen. Due to this characteristic, cancellable biometrics utilizes irreversible transformations or salting to overcome this serious drawback. Based on the above, this paper evaluates the two cancellable biometric schemes for authentication and identification of users using iris images. Specifically, Gray-combo and Bin-salt techniques are implemented and evaluated according to the ISO/IEC 24745 standard, which considers 8 criteria to estimate the performance, robustness, compliance of properties, and vulnerability of each proposal. The results show that these two techniques have a good performance, privacy and security, allowing their easy incorporation into any authentication or identification system based on iris recognition.