Intelligent surveillance system for event detection

This project aims to develop an intelligent surveillance system for automatic event detection. The proposed system will operate in an indoor environment to notify the user of events of interest in real-time. Most standard systems use visible-light cameras and basic change detection methods (e.g. Background subtraction) to recognize simple events such as intrusion. Instead, we aim to analyze and understand complex real-life activities, which is a very challenging task due to the difficulty of analyzing a 3D scene projected on bi-dimensional images. The novelty of our approach lies in the efficient exploitation of the 3D information obtained from a depth sensor. Our idea is motivated by the recent emergence of the low-cost RGB-D (Red, Green, Blue, -Depth) cameras, providing the depth information along with color images. The project will provide the partner company with a highly competitive product on the cutting edge of the surveillance technology

Faculty Supervisor:

Rita Noumeir

Student:

Wassim Bouachir

Partner:

Aerosystems International Inc.

Discipline:

Computer science

Sector:

Aerospace and defense

University:

Program: