PROJECT TITLE :

Iris Recognition Based on Human-Interpretable Features

ABSTRACT:

The iris may be a stable biometric trait that has been widely used for human recognition in numerous applications. But, deployment of iris recognition in forensic applications has not been reported. A primary reason is the shortage of human-friendly techniques for iris comparison. To more promote the utilization of iris recognition in forensics, the similarity between irises should be created visualizable and interpretable. Recently, an individual's-in-the-loop iris recognition system was developed, primarily based on detecting and matching iris crypts. Building on this framework, we tend to propose a replacement approach for detecting and matching iris crypts automatically. Our detection methodology is in a position to capture iris crypts of various sizes. Our matching scheme is meant to handle potential topological changes within the detection of the same crypt in different pictures. Our approach outperforms the known visible-feature-primarily based iris recognition method on three completely different data sets. In explicit, our approach achieves over twenty two% higher rank one hit rate in identification, and over fifty one% lower equal error rate in verification. Additionally, the good thing about our approach on multi-enrollment is experimentally demonstrated.