Embedded Intelligent Home Network Video Surveillance Based On ARM And Linux

With the rapid development of electronic information technology and network technology,science and technology is changing the way of people’s life step by step.smart home also came into being,it makes people’s lives more secure and comfortable.Security system which is the most important part of Smart home,has attracted more and more attention and come into people life.Embedded video surveillance has become the main force in the field of intelligent home video surveillance with its stable performance,compact size,low price etc.The design uses ARM9 architecture based on S3C2440 A high-performance low-power processor as the system’s embedded processor.Using a USB camera to capture video data,Web server is transplanted in the system,so that users can remotely through the Web browser to monitor the image of the camera view.In the aspect of innovation,the detection of moving objects is added to the system,and the improvement is made based on the traditional Gaussian mixture model,which makes the detection of moving objects more rapid and accurate.Meanwhile,the GPRS alarm system is added.When detecting moving objects,the system automatically send alarm message to the user.The system realizes the video monitoring network,remote,intelligent,but also effectively reduce the cost.In this paper,first introduces the smart home,video monitoring and embedded system,according to the current development of the video monitoring to determine the overall function of the design,based on the analyzes and introduces of the current embedded processor,the system choose S3C2440 A chip which based on ARM9 architecture as the system’s embedded processor.The system hardware platform is composed with the external storage part,the network part,GPRS alarm part and so on.Then,the current mainstream embedded system is introduced.According to the difficulty of development and cost,the Linux system is selected as the embedded operating system,and the Bootloader,Linux kernel and root file system are implemented on the hardware platform.Compiled and ported.In the detection of moving objects,after compared with the optical flow method,frame difference method,background subtraction three detection techniques finally selected background subtraction,through improved Gaussian mixture model to achieve the detection of moving targets.Finally,the system uses Boa server,through the B / S mode so that users can realize remote monitoring purpose.After the system is started the user can view the video surveillance screen through a networked computer Web browser.When the system detects a moving object,it will automatically send text messages to User and remind the user to view the surveillance video.