Traffic queue length measurement by using combined methods of Photogrammetry and digital image processing

Real-time measurement of traffic parameters is increasingly demanded for monitoring the traffic status and adjusting the timings of traffic lights. In this paper, we propose an economic and operable approach to measure a traffic queue length on real-time by employing the combination of photogrammetry and image processing methods. The requirement for our development is rather simple: one camera which is fixed in a proper place and a 2D digital vector map. Therefore, it is fairly economic and feasible. There are four steps needed to acquire traffic queue length: control points are derived from a 2D digital vector-map; the interior orientation (IO) and exterior orientation (EO) parameters are extracted; the image is processed and coordinates of vehicle headlights are calculated; vehicles are separated into each lane and the length of traffic queue is calculated. Steps and are only implemented once when a camera is installed. After the parameters of IO and EO are obtained, they can be applied repeatedly. Therefore, this method can be operative in real-time. In addition, separating vehicles into each lane has been a challenging work in the traffic queue study. Our method has solved this problem and achieved a promising result.