Active Object Tracking based on Pan-tilt-zoom (PTZ) Camera for Smart Surveillance System
Hwiseok Yang, Yoo-Joo Choi · Jeongbo gwahaghoe nonmunji. so'peuteuweeo mich eung'yong · 2011
In this paper, we propose a real-time active tracking method based on an adaptive background modeling for an environment to use PTZ network cameras. The proposed method effectively removes a ghost object and robustly tracks a stopped object by applying a 2-level background updating scheme based on the adaptive media filtering background subtraction and motion history analysis. For robust active tracking to a target object, we apply an estimation algorithm of the camera FOV change position which considers the position and velocity of a target moving object, delay times in networking and PTZ camera control. Furthermore, we also present a method to control zooming parameter of the PTZ camera in order to magnify the target object in the center of camera FOV. In the experiment, we show that the propose method is effective in ghost removing and tracking of a stopped object, and prove that a target object can be robustly tracked after Panning/Tilting/Zooming events in real-time.