The Microscopic Traffic Flow Information Collection And Analysis In Roundabouts With Traffic Signal

The different geometric of roundabouts' s ring with traffic signals decide different operational characteristics of vehicles.Because roundabout's vehicles with traffic signals are running complex,it is especially important to driver behavior identification on traffic information collecting and using this information.Most parameters of traffic flow collection device is mainly: macro parameters such as,flow rate,density and speed section,headway,These parameters do not meet traffic flow theory and simulation study on microscopic traffic flow needs.Microscopic traffic flow information is the parameters information about traffic individuals within a certain time,space,continuous trajectory and speed,acceleration and so on.Gets the microscopic traffic flow parameters is the basis and key what carried out research and microcosmic simulation of traffic flow theory,based on image processing method of microscopic traffic flow is a protection what gets the microscopic traffic flow parameters.Image processing technology is a microscopic traffic flow data collection methods,compared with other traffic information collection,It uses advanced image processing,computer vision,and pattern recognition technology,to a certain extent greatly saves manpower and material resources,while also reducing the cost of scientific research and the data easy to store.This paper based on the video of signal controlled roundabouts,using image processing technology,through video processing data for the vehicle trajectory,getting traffic microscopic traffic flow information for individuals on the road,and analyzing the characteristics of driver on credit control-loop,lay the Foundation for traffic simulation model and parameter calibration,main tasks as follows:(1)To the captured video conduct pre-processing through uses image grayscale,image filtering,image enhancement techniques,and solving method and experimental results are given.(2)Details of camera model and comparative analysis of two kinds of traditional camera calibration techniques,choosing the right camera calibration technique to achieve the conversion between image and actual,and through two different method to make camera calibration on signal-controlled roundabout.(3)Using open-TLD based on TLD tools to vehicle which on the signal controlledroundabouts video detection and tracking,get vehicle image coordinates information on signal controlled roundabouts,and evaluate the track results.(4)Using the vehicle's actual coordinates,get track information for vehicles within a certain time period,using tracked vehicles the geometric center of the window to get the vehicle in all moments of velocity,acceleration and the same road number.(5)Using the method which based on image processing on microscopic traffic flow information collection to Explore actual different flow vehicle's velocity and acceleration and exchange rate variation on signal controlled roundabouts circulation area,interweaving,and make a analysis to signal-controlled roundabout driving behavior.