Research And Implementation Of Intelligent Analysis System For Highway Video Surveillance Based On Cloud Platform

In recent years,along with the development of industry of our country road traffic is becoming more and more rapidly,highway travel demand continues to increase,the existing expressway video monitoring system of bearing pressure,and various provinces and cities construction there are many differences in video monitoring system,so in the nationwide,is difficult to realize video monitoring system of the in-depth application of networking and intelligent analysis technology.The disadvantages of the traditional highway video surveillance system can not meet the increasing needs of the road network center for the overall supervision of the highway and the rapid disposal of emergencies.The expressway system urgently needs a new type of video surveillance system architecture,in order to build a highly unified national standard,and can realize the full network of video resources and video intelligent analysis of the expressway video surveillance system.With the increasing development of cloud computing technology and deep learning technology,it is possible to use advanced cloud platform technology and intelligent detection technology on the basis of expressway video surveillance business to make the expressway video surveillance system have the ability of "mass fusion,automatic identification,emergency analysis and intelligent control".This paper discusses a method to transfer the existing expressway video monitoring resources to cloud platform.Firstly,cloud computing resources are established at the ministerial and provincial levels,and a set of expressway video monitoring system based on cloud platform is deployed on this basis.In addition,cloud gateway and other cloud video devices are deployed in the section monitoring sub-center.The upper cloud gateway accesses the underlying video monitoring resources and completes the registration in the upper system.At this point,the upper system can make the upper cloud gateway push the specified video stream to the resources of the fair station through instructions.After the system completes relevant services such as video management or video storage,the video resources can be pushed to all network users through the efficient content distribution network of the cloud platform.Through this system,the video monitoring resources of the road section can be unified and collected on the cloud,and the video monitoring resources can be flexibly called according to the business needs.In this paper,the highway intelligent detection technology was studied,to build a target detection based on freeway scene data sets,and on the basis of Faster-RCNN network model,by the highway scene training data set to optimize the network parameters,and to improve the detection ability to increase the data augmented model,improve the detection effect of network model.Then,based on the highway video monitoring system based on cloud platform,the optimized intelligent detection module is integrated into the system.In this way,instead of the traditional human eye observation method,the intelligent detection technology is used to achieve the purpose of video real-time monitoring of the road section.In this paper,the final design and implementation based on the cloud platform of highway intelligent analysis system,video monitoring in massive video surveillance resources together with cloud technology network at the same time,in depth study of intelligent detection technology of video monitoring intelligent degree,for the highway supervision system has very important practical application value.