The U.S. Army maintains numerous forward operating bases (FOBs) that each require strict security. These FOBs are monitored by a collection of cameras constantly producing live video feeds of the surrounding area. If automated algorithms could detect and track people in each cameraâs field of view (FOV), security could be improved by identifying threatening individuals, preventing infiltration by non-authorized personnel, and rapidly authenticating authorized personnel. The technologies to support this use case are emerging but they are disparate and do not natively scale to scenarios of this scope. What is needed is a multi-tiered system that connects to Army assets, analyzes live video, and scales to dozens, hundreds, and thousands of cameras or more. In this project, we will leverage technology and experience from ourselves and our partners, including person detection and tracking, facial recognition, edge computing, image enhancement, real-time processing, and integration with Army systems to develop a scalable solution for performing live biometrics analysis from Army cameras
