Iris recognition system with error detection and reconstruction algorithms for template security
Kumari A Anitha, Maya V Karki · 2017
The essential concern related to biometric structures which are secured in a database is the security of the database storage system. Most of the existing algorithms need to decrypt the whole dataset before matching of query iris image that is a time-consuming, challenging and tedious mission. It also affects the performance of various parameters of the framework. If the encrypted dataset is damaged the whole system fails. In order to overcome these issues a new technique which gives security to an iris feature without affecting the performance of the whole system is implemented using fallacy clear-up algorithms. It includes error detection and reconstruction of damaged templates. In this paper, three different algorithms for error detection and reconstruction on the binary iris template are proposed. The memory requirement and recognition rate of the system with and without the fallacy clear-up algorithms are compared and also the three different fallacy clear algorithms are compared with respect to computations, time and memory requirements. Finally, an iris recognition system with automatic error detection and reconstruction process is designed and tested using an UBIRIS iris database. The proposed algorithm detects an error and reconstructs the damaged template before every query with the recognition rate of 99% for UBIRIS and 100% accuracy in error detection.