Research And Implementation Of Smart Home Monitoring System Fused Intrusion Detection

With the development of economy and society,people began to pursue a safer,more comfortable and more efficient living environment,and an intelligent home life gradually entered into people's daily lives.At present,smart home systems still have deficiencies such as high prices,incomplete functions,and poor security in video monitor.This article focuses on the above deficiencies,focuses on the relevant visual detection algorithms for external intruders,based on an Cortex-A9 architecture ARM chip equipped with an embedded Linux operating system,combines key technologies such as ZigBee wireless communication,designed and implemented a smart home monitoring system that integrates detection of intruders in the interior and complete functions.The system is divided into two parts: hardware and software.The system hardware development platform is divided into three parts: embedded gateway module,ZigBee wireless communication module,Image acquisition and alarm module.The embedded gateway module adopts the Cortex-A9 architecture of the Samsung Exynos 4412 as he core processor for the development board,and the ZigBee wireless communication module selects the CC2530 microcontroller.Smart home monitoring system software design includes development environment setup,video image acquisition design,communication protocol design,serial communication program design,embedded gateway server design,and Qt client and web client design.Video image acquisition adopts V4L2 architecture to achieve the acquisition of home video image information;communication protocol design is the rule that stipulates the data exchange between related modules;serial communication program design realizes the communication function between ZigBee coordinator,GPRS module and embedded gateway;The design of server and client is based on two modes of C/S and B/S,which realize the interaction function between smart home monitoring system and users.The design of intrusion detection alarm module mainly improves the shortcomings of the intrusion detection algorithm in the current smart home video monitor system.By transplanting OpenCV in the embedded Linux platform and combining the OpenCV technology,first to detect the presence of a moving target in the video image,after finding a moving target,use the HOG+SVM machine learning method to determine whether it is a person.If the discriminate is human,the system will automatically notify the user via SMS via GPRS module.Finally,the system is tested.The experimental results show that the system meets the established requirements and has good stability and accuracy.