| | |
Summary: Video Surveillance of High Security Facilities
S. Kang*, A. Koschan, B. Abidi, and M. Abidi
Imaging, Robotics, and Intelligent Systems Laboratory,
The University of Tennessee, Knoxville, TN, USA
*sangkyu@utk.edu
Abstract An integrated system for automatic detection, handover and tracking of an intruder using one fixed camera
and one Pan/Tilt/Zoom(PTZ) camera is presented. This system includes automatic detection using an overhead static camera
of an intruder who walks into a crowded secure area, intruder handover from the static to the nearest PTZ camera, and
tracking using the PTZ camera. The intruder is detected by applying optical flow algorithms to video from the overhead
camera. A PTZ camera extracts the intruder detected by the overhead camera and builds a color histogram model of the
target. A novel region segmentation based on the color histogram and the algorithm for tracking are presented with real
experimental results.
I. INTRODUCTION
The desire for safety in both public locations where
many people gather and high security facilities, such as
nuclear plants and research laboratories, has been
increasing since September 11. For security purposes,
such locations utilize biometric-based recognition tools
for access control as well as video camera-based
surveillance systems for monitoring activities. Finger
|
