The third and final Video Surveillance
and Monitoring (VSAM) Demonstration will be held on October 19, 1999 at
Carnegie Mellon University, Pittsburgh, Pennsylvania.
Presentations will
cover the underlying automated video understanding technologies for moving
object detection, tracking, and classification; multi-camera coordination;
real-time geopositioning ; VSAM system architecture; the operator control
interface; and environmental modeling for video applications.
To facilitate
technology transfer and minimize cost, attendence has been limited
to Government representatives, including a broad spectrum of technologists
from OSD and the Service labs formulating a new OSD initiative entitled
Smart SensorWeb.
VSAM research is sponsored by the Defense Advanced Research Projects Agency
(DARPA) as one of three battlefield awareness initiatives under the Image
Understanding Program. The objective has been to develop automated video
understanding technology within networks of "smart" cooperating video
sensors. Potential applications use in future urban and battlefield
surveillance range from building and parking lot security to monitoring
restricted access areas in warehouses and airports, scanning urban
environments for sniper activity, and performing reconnaissance on the
battlefield. Technology advances developed under this project will enable
one human operator at a remote workstation to supervise a network of VSAM
platforms (stationary, moving on the ground, or airborne), having multiple,
steerable sensors operating in the visible and infrared bands for continuous
day/night operations. The platforms will be largely autonomous, notifying
the operator of salient information as it occurs. In addition to advanced
video/image processing and compression techniques, a great deal of
collateral data and contextual knowledge will be incorporated into the VSAM
system so that it can reliably perform its assigned tasks.
For more information, contact Stephanie Riso at riso@andrew.cmu.edu by October 13.