Design and realization of intelligent monitoring system based object identification

Video based monitoring system plays an important role in asset management. However, transmission of video data occupies large amount of bandwidth, especially in wireless communication environment. On the same time, searches and storage of video data is a troublesome problem in after-event investigation. This paper studies a non-video monitoring system which is based on RFID and sensors. The system collects identity information and environment parameter information with the help of RFID reader, and then judges event type by the proposed event type deducing model. Every event is record to event log database and illegal event will trigger an alarm signal to watcher and then assets are monitored. Experiment results are the following. First, on average, occupied bandwidth is only 25.0 Kbps when monitoring 1,000 assets. So the system is suitable for wireless environment. Secondly, searches can be done within few seconds accurately benefiting from non-video event log, and efficiency of after-event investigation is improved greatly. Thirdly, storage of 1,000 event log only occupies no more than 300 KB. Fourthly, different monitoring area are communicated by wireless channel which complying with 802.11 b/g, and the whole monitoring area connects with Internet through Root station of the system. Monitoring center deployed in Internet can hold personnel, assets and environment information of the whole monitoring area in real time.