Biometric Feature Extraction in a Multi-Camera Surveillance System

Advanced, multi-camera surveillance systems present a practical and effective means of automatically detecting, tracking, and characterizing various people, events, activities, and situations. Recently, considerable focus has been placed on the tracking and analysis of human motion and, specifically, on the measurement of diverse biometrics for authentication and identification [1]. In this article we concentrate on the use of gait as a biometric and provide an introduction to a unique system for its robust tracking, extraction, and subsequent analysis.