Online Traffic Density Estimation and Vehicle Classification Management System

Abstract

This paper shows the design of an online web based traffic management system that presents the traffic density of heavy and light weight vehicles. The methodology used in this system performs object segmentation using edge detection to identify the object in video frames. Moreover, the robust object identification using feature extraction will identify the object in bad conditions i.e. night vision, shadow in daylight and occlusions occurring due to absence of gap present between two objects. The Blob analysis is the part of implementation to identify the type of vehicle (light and heavy) along with calculating the speed of vehicle. Further we are counting number of objects passed using Motion History Images (MHI) which generates the history of each object and assigns label to each object. This paper addresses the issue that there are various paths to same destination in which some of them are densely populated and others are vacant which are unknown to the travelers. Our online system traces the estimated density of heavy and light weighted vehicles of different locations at different times which will certainly help user to decide the suitable path to respective destination. Moreover, User can estimate time that it will take to reach the destination from particular path.