Research And Implementation Of The Intelligent Monitoring System Based On Rfid And Image Understanding

The traditional remote monitoring system is in short of effective video detection and analysis mechanism. In large-scale monitoring systems which monitoring nodes are intensive, considerable amount of labor is needed to ensure that monitoring information is effectively processed. To taff, a long time watching the monitor screen can lead to fatigue, which some important details may be missed. Monitoring information kept by the traditional remote monitoring system is lack of effective indexing mechanism. Besides the time sequence, monitoring information kept by the traditional remote monitoring system is in the lack of effective indexing mechanism. In many cases, it will cost a lot of time and labor to index useful monitoring information from enormous video database just according the time clue. Therefore, it has become an overall trend of monitoring technological development to improve the function, increase system intelligence and work efficiency, and expand the range of system applications.After a detailed analysis of trends and design requirements of the intelligent monitoring system, this thesis has developed an intelligent monitoring system based on radio frequency identification and image understanding, which used existing technologies such as RFID, serial communication, video compression, video transmission, motion detection, target tracking, and relational database. The work and achievements are as follows:Firstly, the read of e-tag information has been accomplished on the basic platform system of XCRF-500. The transmission of the e-tag information has been realized under the serial communication technology.Secondly, the transmission of video information has been achieved by video compression technology such as MPEG2/MPEG4/H.264and RTP/RTCP.Thirdly, technologies such as motion detection, target tracking, behavioral analysis, and feature extraction have been accomplished. In order to improve the robustness of the tracking algorithm, the situation of mutual occlusion of a moving target was considered. Fourthly, SQL Server relational database technology is adopted to achieve video surveillance information storage database. Video surveillance site is built under the Win32system and the. Net Framework environment, which provides users with the query, display, playback and other functions.