Using Image Processing for Microsimulation of Freeways in Developing Countries
Mohammad Sadat Hoseini · Transportation Research Board 85th Annual MeetingTransportation Research Board · 2006
There are many reported researches considering movement behavior of vehicles. In these studies the two basic types of movement of a vehicle are longitudinal movement, car following, and lateral movement, lane changing. Dependency between car following and lane changing models and the fact that many drivers do not consider driving lanes in developing countries, necessitates developing models that combine car following and lane changing models without considering the driving lanes. An image processing system has been invented for determining the position of vehicles in the images taken from the motorway. This data is used for calibrating and evaluating the proposed model. In this research a cellular automata model has been used. In this model a basic freeway segment has been divided into cells. Individual vehicles in each simulation time step occupy single cells and their movements in these cells describe movement of vehicles in the motorway. Microscopic simulation models considering the movement of individual vehicles in a traffic stream can be particularly useful for traffic analysis. This paper describes research to develop a microscopic model for driving behavior based on the optimization process of each driver for increasing his/her speed and decreasing his/her risk of collision. A new concept for measuring the risk of collision is proposed and maximum available speed for the vehicles is calculated using movement regulations.