Intelligent Cooperative Tracking in Multi-camera Systems

In this paper, an approach for intelligent integration of indoor visual tracking system for event detection and movement is proposed. This surveillance system is composed of a stationary camera and a pan tilt zoom (PTZ) camera, where the two cameras have been intrinsically and extrinsically calibrated. The stationary camera detects events such as fall and wandering using motion-based visual tracking. In this initial study, the PTZ camera tracks and follows the person who triggered the event using intelligent color-based particle filtering which is defined based on the expected dynamics of the scene. The purpose of tracking in view of the PTZ camera is to continuously keep the person in the full view of the camera which can further be processed for identifying details of the person. Preliminary experimental results for camera calibration, event detection, and human tracking are presented to demonstrate the performance of the proposed cooperative hybrid visual tracking system.