Remote Intelligent Video Surveillance System Of Coalbed Methane Field

It is extremely important to monitor the coal bed gas field because there is a certain risk that coal bed gas is a kind of inflammable and explosive gas. However, most of the coal bed gas wells are in the remote wilderness, facing the difficult problem of inspection, high wiring costs. Aiming at these problems, this paper designs a remote intelligent video monitoring system for industrial coal bed gas, which is a good solution to the problems above.As for algorithm, after comparing the classical foreground detection algorithm and screening the accuracy and real-time performance, this paper finally improves the ViBe algorithm which was proposed in recent years. ViBe algorithm was proposed in 2009 the first time, also that was the first time to apply the theory of clustering to the foreground detection algorithm, which has the characteristics of little amount of calculation, high accuracy and strong adaptability. At the same time, the ViBe algorithm also has its limitations, such as not adapted to the dynamic background, poor anti-interference ability, prone to "ghost" etc. this paper also mends these problems. Through expanding the selection scope of the background model reasonably and hierarchical deciding the decision rules, the improved algorithm is well adapted to the field of complex dynamic background and can meet the actual testing requirements of the scene. Also the simulation verification has been conducted on the improved algorithm and this paper compares this algorithm with other classical algorithms in terms of accuracy, real-time performance and so on.After the algorithm completed, this paper details the design principles of the client hardware as well as each part including infrared camera, video decoder, ARM controller Cortex-A9, network controller DM9621ANP, industrial wireless router and power supply system, etc. In addition, the design flow chart of the program in the logic controller is also introduced in detail. When the program is running in the hardware, the client device can respond to the abnormal situation, take photos and upload according to prior analysis of the scene. For data transmission, the system build VPN channel multi point to one point using the communication base station and realizes the remote wireless transmission of data.At the server side, this paper designs the monitoring software for data reception, display and control commands send. This software has a series of functions including login, registration, playback, alarm. Also the software can be achieved on the client's remote command control, with a good human-computer interaction interface.At last, the whole system is tested in many aspects on the scene, including accuracy test, real-time test, stability test, etc. Moreover, this paper points out the aspects that the system needs to continue to study and improve in the future along with other fields that the system can be extended to, such as intelligent transportation, production safety and so on.