Design And Implementation Of Remote Video Monitor System Based On Embedded Linux

In recent years, with the development of network communication technology, computer technology, multimedia technology and artificial intelligence technology, people’s demands for safety system are also rising. Video monitoring system has entered a time of being digital, intelligent and networked, the embedded video monitoring technology has become a hot research of security field. For the above conclusions, a remote video surveillance system is proposed in this paper. The system is based on embedded Linux and ARM11 is used as a core component of the design.The system need to realize the following functions:Firstly, the system acquires video images by camera, and secondly, the image data is transferred from server to client through network. Finally, the client displays the video, detects the moving target and realizes the alarm function. Based on the analysis of the system requirements, this paper gives a whole design of the remote surveillance system. On the hardware side, The hardware composing and construction of the development platform, hardware architecture of ARM11 development board and the design of hardware interface circuits are proposed in detail. On the software side, it can be divided into two phases:designing the embedded system and designing application. The embedded system includes Bootloader, Linux kernel driver, device drivers and so on, it provides a running platform for the next development of application. The application has two modes:C/S mode and B/S mode. While in C/S mode,it is divided into two parts:the server and the client. The server is used to offer data stream service by transplant MJPG-streamer server, The video data acquisition can be done by call V4L2 and achieve data transmission by using the tcp/ip protocol. The client receives video data, detects the moving target and displays the image ands the QT development tools is applied to compile the program. The B/S mode is designed based on the server of the C/S mode, and then it can broadcast directly in the browser by transplant the embedded web server.By analyze the demand about the detect moving target function, an improve algorithm for moving object detection based on Vibe is presented.. Firstly, The algorithm get the binary image of pre-detection moving target by the image processing algorithm such as the frame differential method and morphological dilation. Secondly, it combines the Vibe algorithm to update the background mode and detect the moving target, then check the video by comparing the proportion of moving target in picture. it is accomplished by applying computer vision library OpenCV and makes the system more intelligent.The function and performance of whole system are debugged and tested after completing the design of system, the test results show that it is stable and reliable, it meets the requirement on the remote video surveillance system. The system of achieving better performance detecting moving target functional part, thus realized the goal of anticipated design.