An Optimized Iris Recognition System for Multi-level Security Applications

This paper proposes an optimized iris recognition system which is suitable for client applications with several security levels. The innovations are the low dimensionality of the iris feature space and the hierarchical classification-based identification method; in this approach the biometric samples are processed in 3 stages to provide identification final decisions. The proposed design solution provides scalability, flexibility and a low computational complexity in biometric data processing. These benefits are important for costs optimization in security systems design.