Threat-based evaluation for context-aware multimedia surveillance system

Multimedia surveillance has become commonplace and getting increasingly intelligent in recent years. The technological advances in computer-vision techniques help to detect suspicious movement and identify objects. Meanwhile, as surveillance system is getting more sophisticated, the risk of privacy intrusion also increases. Various researches regards privacy protection have been done in multimedia surveillance, however, the implementations in current security systems is limited compared with the surveillance technology. In this paper, a context-aware multimedia surveillance system with threat-based service selection method is proposed. We use pre-defined context to recognize activities and identify different behaviors from multimedia streams. The Analytic Hierarchy Process approach is applied to help user to analyze the services. A user preferred multimedia surveillance service which achieves a desired trade-off between security, privacy and cost is provided automatically.