Design And Implementation Of Multiple Information Source Monitoring System Based On WebRTC

With the increasing complexity of outdoor scenes,people can no longer perceive the detailed situation of the external field through traditional video surveillance.People more urgently need a multi-source monitoring system that integrates video,audio and environmental information.Therefore,in combination with the existing network environment,new technology is applied at the application layer to develop a monitoring system that can provide multiple information sources,which has become an urgent need for application scenarios such as highways and stadiums.In view of the above problems,this paper designs a multi-information source monitoring system based on Google's open source Web RTC audio and video processing framework.The system is divided into four major modules: collection terminal node,primary server,data storage and secondary server.It realizes five functions of user management,information collection,information playback,equipment management and remote control.This article focuses on the following two aspects:(1)Multi-information source monitoring system software implementation based on WebRTC.This article first uses the ARM architecture-based collection terminal node to periodically collect environmental information,transmits it to the server through the TCP protocol,and stores it in the MySQL database.At the same time,it uses Web RTC technology to collect audio and video data,and transmits it to the Kurento server.Format storage,used to support video playback.Then designed the data storage center,encapsulated the database middle layer.Then the web server application program is designed using Go language to provide background support for the front-end interface.(2)Optimization and testing of multi-information source monitoring system based on WebRTC.Firstly,a fair competition model of network bandwidth based on packet group filter feedback is proposed to optimize the network bandwidth ratio of multiple information source systems.According to the linear regression,the frame rate in the packet group is predicted,and the feedback information is used to actively adjust the transmission rate of the two data frames in the network.After the actual scenario test,it is verified that this algorithm can significantly improve the data delay problem of the weaker party in the network bandwidth competition.This model has nothing to do with the data frame type and network parameters,and is universal.This system uses black box testing and integrated testing methods.It has done a lot of experiments in system functions,real-time performance and stability,and has verified that this system can meet the user's functional and performance requirements.At present,the actual project designed in this thesis has been completed.After site investigation and user feedback,Party A is basically satisfied with the system's interface,functions,and performance,and the system runs stably.