Design And Implementation Of Remote Monitoring Video Preptocessing Terminal Based On Embedded Vision

At present,the remote intelligent video surveillance system handles the surveillance video mainly in the centralized processing mode of the monitoring center,which requires the monitoring center to process a large number of surveillance videos containing normal environments.When the number of cameras required for monitoring is large,centralized processing will bring large amount of computation to the image processing server of the monitoring center.Therefor,a distributed monitoring video preprocessing terminal system based on embedded visual technology is designed and implemented in this thesis.The terminal system identifies the abnormal video in the monitoring area,does not upload the normal video,only uploads the uncertain and suspected video,which can effectively reduce the workload of the monitoring center image processing server and also reduce the transmission of monitoring video data.Linux embedded operating system technology,V4L2(Video For Linux 2)video image acquisition technology,digital image processing technology,wireless 4G communication technology,step motor control technology are used by this thesis,to design a remote monitoring video distributed preprocessing terminal system based on embedded vision.The system includes video and image acquisition unit,main control and digital image processing unit,Ethernet transfer 4G data transmission unit and cloud platform control unit.The video image acquisition unit uses ZC301 USB camera and V4L2 video image acquisition technology to complete the collection of unattended surveillance area images.The main control and digital image processing unit adopts S3C2440 processor based on ARM920 T to perform background modeling,foreground extraction,initial image recognition and other operations on the collected video images,processing the communication data,and generating the signals needed to control the step motor of the cloud platform.The Ethernet transfer 4G data transmission unit combines the Ethernet module on the terminal circuit board and the 4G router to transmit the abnormal message data and the abnormal video image data from the main control and digital image processing unit to the monitoring center through the Internet,and responsible for receiving the response,command and other data transmitted by the monitoring center,and transmitting it to the master and digital image processing unit.The cloud platform control unit control the stepping motor to rotate according to the control signal the main control and digital image processing unit sends,to achieve the control of the cloud platform of the camera.After the module test and thelaboratory simulation debugging,the terminal designed by the thesis can accurately judge the abnormal situation and automatically upload the abnormal video,and can also detect the abnormal situation of the suspected flame,smoke and pedestrians with certain probability,to share the preprocessing work of the monitoring center,to lay the foundation for further identification of the monitoring center and reduced the processing pressure of the monitoring center.The system is flexible,reliable,simple in structure,easy to maintain,and low cost.The design goal of reducing the pressure of monitoring center and reducing the transmission amount of video data is achieved.It provides a certain reference for further research on embedded distributed preprocessing technology.