| Currently,video monitoring system has been widely used in every aspect of life and is playing an essential role.Hohhot Railway Bureau also deployed the video monitoring system in its communication equipment room,together with the supervisory system of power &environment,to serve for the environmental monitoring,security-monitoring and accident investigation and evidence collection.However,there are some flaws in the current video monitoring system of the unmanned computer rooms along the Railway Bureau.First,the long-time and uninterrupted video collection requires the very high performance of storage equipment and often causes heavy loss of memory disks.Second,video storing demands a high level of transmission bandwidth,thus occupying most of the transmission resources.Third,the redundant information in the video makes the investigation and evidence collection of the accident highly time-consuming and strenuous.After a careful study and analysis,combined with the status and working characteristics of the unmanned computer room along the railway,two requirements are put forward for the video monitoring system.First,it should monitor the computer room environment in real time and ensure that maintenance personnel can access to the image and working status of the computer room at any time.Second,it should store the accident image in case of any abnormality in the computer room,so that to facilitate the investigation and evidence collection process and accident analysis and responsibility identification.Based on the above-mentioned needs,this paper designs a video monitoring system with ARM9 microprocessor,embedded operating system of Linux and OpenCV computer vision technology as its core.Through real-time target tracking and dynamic calibration algorithms,the video image data can be stored only when there are moving objects,which is more concise,effective and storage-and transmission bandwidth-saving.At the same time,after taking the stability,maintainability,portability of the system and cost performance of the equipment into consideration,the paper has done a lot of work on hardware selection and system compatibility.the main jobs and results are as follows:1.A suitable embedded system architecture with ARM9/S3C2440 microprocessor module as its core was designed,together with the power supply module,USB camera,network port communication module,SD memory card and other functional modules,to realize the functions of the video monitoring system.2.After examining the existing video monitoring system on the market,a set of highly efficient,reliable and cost-effective hardware was selected and designed,and circuit design is carried out fully considering the interference between the modules and the compatibility of the system.3.The main program workflow for the video monitoring system are designed and the implementation of dynamic detection algorithm is optimized.Finally,after the functional design required is completed,the equipment debugging is conducted to verify the system and results are as follows: all the hardware functional modules of the system can work normally and that the video capture has good real-time performance without obvious time lag;the effect of image acquisition and display in the real-time interaction process meets the standard;in the monitoring target area,the video image data can be saved only when there is moving object and it is marked so as to facilitate tracking process.In this way,the storage loss,image processing time and transmission bandwidth of video monitoring system are reduced significantly. |
PDF Full Text Request