Unusual activity detection for persistent target surveillance
Zhihai He · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2005
The information generated by the integrated sensor suite is massive. Simply bringing the information to the decision makers is a cognitive disaster due to the information overload. Intelligent algorithms need to be developed to exploit and filter the information. In this work, we seek to develop content-level change detection and activity-level information fusion and filtering algorithms to detect unusual activities for persistent target surveillance applications.