Design And Implementation Of Intelligent Monitoring System Of Regional Intrusion Detection Based On Video

With the growing of social population and expanding of city size, people’s requirement to safe environment through management is becoming higher, especially for some important occasions whose monitoring requirement is high like prison and bank, and for some complicated fields whose monitoring difficulty is critical like wild camp. Traditional video monitoring can’t fit people’s requirement now, it’s very important and necessary to monitor these sensitive areas with efficient intrusion detection methods which can give pre-warning timely when dangerous actions happen, in order to remind workers.This paper mainly study the target tracking and recognition of intelligent monitoring system. Firstly this paper use characteristic matrix to track multiple targets based on kalman tracking which can solve the problems that multiple targets interfere with each other and tracking precision is low, secondly propose a characteristic distance model based on computer graphics to solve the problem that it’s difficult to recognize target action under complicated background, and take the fault tolerant strategy to cope with problems brought by different point of view of cameras, finally use Open CV2010 and Microsoft Visual2010 to design and realize this intelligent monitoring system. The system is divided to two parts, i.e. analyzer and client, to ensure data can be processed synchronously and in real time, and customer can carry and use expediently.Field test shows that this intelligent monitoring system realized wild filed monitor and pre-warning. Statistics of test cases indicates that the intrusion detection algorithm of this system can reach 88% higher accuracy rate and 2% lower missing report rate, in other words this system has achieved reliable and real-time monitor.