The Design Of Video Monitoring System Based On Embedded ARM9

With the rapid development of embedded technology, image processing and computer network technology, the video monitoring system, as a kind of important life support equipment in our daily life, has become a safeguard system with more advanced and stronger capability. Video monitoring system has a demanding requirement for stability, compatibility and real-time reaction. Meanwhile, according to different scenarios, the specificity and intelligence of video monitoring system have different development directions in many areas of the society.From the family level security, this paper presents a in-depth study to the current situation and features of the video monitoring systems which are applicable to the family plots and the university students’apartments, and then shows a concrete overall design, functional planning and implementation of intelligent video monitoring system with ARM9and embedded Linux technology as the core. The system is mainly using ARM9processor as the core, to achieve the main function of the video monitoring system in the embedded hardware platform through the software implementation. That is, under the unattended condition, through moving target detection and tracking for the fixed monitoring scenes, and judging the size of dynamic object, the system can automatically carry out a real-time alarm, and save the trace of the moving target through the network transmission.The system uses image processing technology to realize the detection and tracking for moving target. As the first step of video processing technology in the intelligent video monitoring system, moving target detection is a hot and difficult issue. Based on the deep study of the moving target detection in a variety of traditional algorithms, the paper uses a target detection method combining methods of frame difference and background difference together. Additionally, self-adaptation mixed Gauss model is also applied in the background updating link, with this method the prospective moving targets can be successfully extracted, which can help to better serve the modeling of moving target tracking algorithm based on regional characteristics of the Kalman filter pre-estimation. At the same time, this paper presents a false alarm processing method for size judgment of dynamic target. Its role is to help the system judge the size of the target according to the threshold value after the detection of moving target, and then decide whether the alarm should be caused. Finally, the system functional testing is presented, the intelligent video monitoring system designed upon the moving target detection and tracking algorithm has high accuracy, and can meet the need of real-time detection and alarm, with a certain practical value.