Measurement of heavy traffic in a tunnel from image sequences
Hideo Kuboyama, Shinji Ozawa · Systems and Computers in Japan · 2004
Abstract The measurement of traffic plays a major role in Intelligent Transport Systems (ITS) in providing beneficial information to the control center or drivers. In this paper, the authors propose a technique for counting the number of vehicles by detecting vehicles from images taken by cameras installed in tunnels. Since the camera height within a tunnel is low, the vehicles overlap when traffic is congested and it is difficult to separate them. As a result, the authors herein propose a powerful measurement technique for congested traffic based on vehicle edge information by using a horizontal edge and the left side of a vehicle to create a vehicle model and performing a cumulative evaluation according to settled frame tracking of this model. In addition, they applied the proposed algorithm to actual in‐tunnel images and showed the results of a vehicle counting experiment. © 2004 Wiley Periodicals, Inc. Syst Comp Jpn, 35(13): 52–62, 2004; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/ scj.10287