The Design Of Embedded Private Residence Security System Based On Machine Vision

With the development of the times,people need more and more security for private houses.Designing a monitoring system based on machine vision intelligent monitoring for personal use has great research significance and application value.It can protect private homes 24 hours without manual monitoring.This topic mainly explores new development directions for exploring the security of private houses.This article will use embedded technology to design a video capture device,and then build a server for communication.The software will use the video capture device to send the captured images to the server.The server processes the server data and sends it to the user mon itoring side.The paper first analyzes the requirements of the system and gives the overall design scheme and ideas of the system.The software design and implementation method of each part will be discussed in modules.In this paper,a background model is established by using a mixed Gaussian background model,and then a background difference method is used to implement dynamic target detection.Then use the HOG feature + SVM to detect the target in the image.At the same time,the background,significance,status quo and development trend of dynamic target detection technology and target recognition technology are also introduced in the article.Then,a combination of target recognition technology and dynamic detection technology is used to design a d etection algorithm for private house security based on machine vision.Finally,the system is used to test the scenario of simulating a private house.The experimental results show that the algorithm and the monitoring system can ensure that the surveillance video can detect the people who invade the private house and report the alarm.Achieve the desired goal.