Design And Implementation Of Remote Real-Time Monitoring Platform Based On Cloud Transfer

In recent years, with the gradual expansion of the scale of do mestic surveillance video, the demand for various types of video s urveillance are endless. With the development of the Internet, remo te real-time video surveillance platform has become a hot spot in t he field of video surveillance. Traditional remote real-time monitori ng platform uses direct network video streams such as H.264 and RTSP streaming media technology,or video frames compressed into frame images, based on Websocket programming and direct trans mission through the Internet. But this way can not always get rid of high-cost remote transmission and poor stability of the problem.The calibration display of counting, intrusion and other video sem antics results also has the problems of weak interactivity and destr uction of the original video. Most of the traditional video surveilla nce platform using C/S architecture, all kinds of development and i nstallation of the client leads to poor versatility and scalability.To solve these problems, we design and implement a remote r eal-time video monitoring platform based on cloud transfer technolo gy. In this platform, we use cloud storage as a remote video data t ransfer channel, forward and process video streams efficiently and i n real-time based on FFmpeg multi-threaded,convert the video strea m to video files and upload them to the cloud, put forward the vid eo backlog processing technology based on heap sorting and adopts Web long polling video real-time acquisition technology,to achieve high real-time and reliability of the remote video transmission tran sfer. In this thesis, B/S architecture and response-based layout techn ology to achieve the PC and the mobile terminal layout and functio n of unity, realize the general display of video semantics by HTM L5 timing superposition technique. In order to evaluate the system,a lot of experiments have been done in the aspects of system relia bility, real-time performance, video quality, response time and sema ntic calibration result display accuracy, which proves that the syste m is efficient and reliable.