Human behavior analysis and its application to a pervasive computing security system

In recent years, the research on intelligent environment based on pervasive/ubiquitous computing technology has attracted more and more attention, especially in security, such as KidsRoom by MIT AI Lab and SmartLab System. In this paper, we illustrate an intelligent security system based on a Human-Centered Pervasive Computing system model which is proposed in the view of engineering. When strange person intrudes into the surveillant environment, this system can detect his activities, analyze his behaviors, and send out alert messages. Six simple human behaviors can be recognized which are walking, standing, sitting, squatting, leaning and lying. We design the security system based on pervasive computing and propose the implementation methods step by step.