Design And Research Of Smart Home Ecology And Security System Based On Cloud Platform

Security monitoring has always been a hot topic of concern at home and abroad.Taking China as an example,the security monitoring system can be seen from the streets and alleys of the city to the advanced conference rooms,hotels,restaurants,and even remote rural areas.However,it is mainly based on the traditional security monitoring.Although the traditional security can record the changes of the scene in the form of video for a period of time,it must rely on human real-time observation to capture the abnormal information of the image.Once there are many monitoring points,it is impossible to monitor all the scenes.Only after the accident occurred,did we go to check the monitoring videos to see what happened,which led to a serious lag in information feedback.In response to the phenomenon,this subject adopts an efficient image detection algorithm,combined with safe cloud computing and high-speed development of Internet communication technology to achieve intelligent monitoring,which makes security monitoring get rid of the dependence on people and make up for the defects of the traditional security system.The system is aimed at smart home security.It's hardware platform adopts the Samsung S5PV210 processor based on the ARM Cortex-A8 core,transplanted open source embedded Linux as operating system,combined with cross-platform Qt technology,and finally completed the power line carrier communication technology for data transmission,USB camera image acquisition and other functions,for image data,this topic uses a compression algorithm to handle remote transmission over Ethernet.The software development is mainly oriented to the cloud server.It not only designed the corresponding data decoding,image processing and MySQL database services,but also built the JavaEE MVC architecture for the better support of web and mobile terminals.In addition,through the research of multiple moving objects detection algorithms and human recognition algorithms,the processing structure of three frame difference method and HOG+SVM human body detection algorithm is improved,the segmentation of moving objects and the discrimination of human characteristics from the image are implemented.Finally,we completed the cloud platform construction and software program development,and on this basis,we tested the system function.By constantly searching for the defects of the system,improving the image processing algorithm and optimizing the multithread data processing,the algorithm has greatly improved the speed of operation and detection precision.It is hoped that the future smart home security monitoring system can be continuously upgraded,not only in the laboratory,but also in the market.Give full play to its practical value.