Combining iris and periocular biometric for matching visible spectrum eye images

This paper describes biometric matching of eye images captured by the visible spectrum smart phone cameras from the MICHE II database. The matching is performed by calculating matching scores for iris codes and periocular biometric based on the Multi-Block Transitional Local Binary Patterns. The authentication scores are calculated separately, and the results are fused to improve the system performance. Score fusion significantly improves error statistics compared to individual iris and periocular matching. For example, equal error rate is about two times lower for fused scores compared to the separate iris and periocular matching. Our matching algorithm ranked first in MICHE II competition.