Embedded Intelligent Video Surveillance System Based On ARM

With the rapid development of computer graphics technology and the continuous optimisation of embedded devices in recent years,a new generation of video surveillance systems has emerged and is widely used in various fields such as home security,campus life,intelligent city traffic and other fields as people's awareness of prevention in modern society increases.The research and development of embedded intelligent video surveillance equipment is favoured by researchers because of its miniaturisation,intelligence,high cost performance and the ability to analyse and process video images.Based on the above background,this thesis designs an embedded intelligent remote video surveillance system,which can achieve the detection and early warning of moving objects.The system is based on a software and hardware development environment combining ARM microprocessor and embedded Linux operating system.The hardware is based on the Exynos 4412 chip,which is an ARM Cortex A9 quad-core processor.In terms of software,the system is developed and ported according to functional modules,a Linux operating system is built on the embedded device,the computer vision Open CV library,the video capture module V4L2,the embedded web server,the embedded email alert module and the moving object detection module are ported.One of the main functions of an intelligent video surveillance system is the detection of moving objects.This paper first introduces three types of traditional motion detection algorithms,followed by an experimental analysis that summarises the advantages and disadvantages of these algorithms.According to the characteristics and limitations of embedded devices,an improved motion detection algorithm combining the three-frame difference method and the Vi Be algorithm is designed,and the detection of moving objects is implemented in combination with the Open CV vision library,and finally an alert email is sent to the surveillance personnel based on the detection results to realise the alert function.The intelligent video surveillance system designed in this thesis is able to achieve real-time monitoring,detection of moving objects in the surveillance area and early warning functions.Experiments show that the system can achieve the basic functions and has a certain degree of intelligence,can effectively alleviate the void phenomenon in the three-frame method and accelerate the elimination of the "ghosting" phenomenon in the Vi Be algorithm,to verify the feasibility of the algorithm,has some practical value.