Distributed Video Analytics and Anomaly Detection
F3-G (Phase 1)

Download the 2013 Project Report

This project investigates the development of automated tools for video monitoring using multiple cameras in support of security applications. The project has both experimental and theoretical components. In Year 5, we focused on two problems of critical operational interest to surveillance and security in mass transit environments: the automatic, real-time detection of counterflow (e.g., passengers moving the wrong direction through a one-way exit lane) and the re-identification of humans in a wide-area network of non-overlapping surveillance cameras.

The goal of this ALERT project is to provide an automated capability for analyzing crowded videos and detecting anomalous motions that can provide indication of activities of interest to human operators in both indoor and outdoor environments.
F3-G Progress Report, Project Overview
Project Leader
  • Richard Radke
    Professor
    Rensselaer Polytechnic Institute
    Email

Faculty and Staff Currently Involved in Project
  • Venkatesh Saligrama
    Associate Professor
    Boston University
    Email

  • Mohamed Elgharib
    Post Doctoral
    Boston University
    Email

Students Currently Involved in Project
  • Ziyan Wu
    Rensselaer Polytechnic Institute
  • Jing Qian
    Boston University
  • Yuting Chen
    Boston University
  • Yang Li
    Rensselaer Polytechnic Institute