Video Surveillance for Biometrics: Long-Range Multi-biometric System

摘要

The human iris is hypothesized to be the best biometric characteristic in terms of uniqueness and robustness. Iris recognition algorithms developed over the last decade have matured significantly to address population-level cross comparisons. Yet iris acquisition systems remain borderline intrusive and less-friendly for subjects and operators. This paper addresses the issue of strictly-constrained iris acquisition in traditional systems. We highlight the observation that all traditional iris recognition systems impose substantial constraints on subject position and motion during iris acquisition. We further observe that the efforts to relax these constraints for iris acquisition of distant and/or moving subjects fall short of scalable system design. We present a novel iris recognition system for long-range human identification. The system is capable of acquiring face and iris images from multiple humans present anywhere in the capture volume. The iris acquisition system uses multiple cameras with hierarchically-ordered fields of views, a highly precise pan-tilt unit (PTU) and a long focal length zoom lens. The system is driven by innovative algorithms that perform wide-area video surveillance, object detection and tracking, and precision pointing. Experimental results are reported in an indoor environment for multiple subject iris recognition at a distance. Eagle-Eyes is a long-range multi-biometric system that improves on existing iris acquisition approaches in terms of stand-off distance and capture volume through the use of collaborative scene and face tracking.