Research On Crowd-sourced Traffic Information Collection And Sharing System Based On Mobile Terminal

The number of private cars has increased drastically due to the improvements in people's livelihood,hence the inconvenience caused by serious traffic congestion.Thanks to the galloping development of information technology and broadband Internet,traffic surveillance,analysis,early warning,forecast and management through Internet of Things and big data are becoming important research projects in the promotion of smart city and smart traffic progress.These research projects remain to be figured out and are current trends in the field of information technology.People have been more dependent on e-maps to check out road conditions because of the strong capability of e-maps.A review of a trail of e-map's functions reveals that using colors,say,red,yellow,and green,to distinguish the level of traffic jam can hardly satisfy people's needs for an explicit knowledge about road conditions.The traditional method of collecting traffic flow needs high cost,so this paper proposes a method based on swarm intelligence to collect and process the data.This thesis intends to conceive a prototype of data collecting and sharing of road conditions on the basis of crowd computing.In addition to visual display of maps,multimedia information about real-time road conditions are collected,analyzed,and screened to achieve the prototype's purpose that the public can gain better knowledge about real-time traffic conditions in a more explicit,more rapid,and cheaper way.The methodology of the thesis is as below:first,smart phone is viewed as mobile sensor;second,cellphone and its camera finish the tasks of collecting photos and describing road conditions.Then,feature extraction in histogram of oriented gradient and the Support Vector Machine theory are employed to identify and screen the vehicles in the images.Deterministic Finite Automata algorithm is used to filter out sensitive words.Also,by virtue of the Game theory,this thesis designs a mixed incentive mechanism which encourages participants' good behavior through incentive measures and credit assessment in order to gain more effective and real information.Last,the efficiency of the whole design is demonstrated by prototype implementation and experimental results.The research in this thesis is advantageous to maintaining urban roads clear and saving time and urban road resources through correct guidance.Furthermore,the capacity to respond to emergency can be improved accordingly.The research can also give a strong impetus to the promotion of smart city and smart traffic progress with the development of data size and road forecast.