Monitoring entire-city traffic using low-resolution web cameras

Abstract

We propose a new approach to intelligent transportation systems for developing countries. Our system consists of two major components: (1) Web-camera-based traffic monitoring and (2) network flow estimation. The traffic monitoring module features a new algorithm for computing the vehicle count and velocity from very low-resolution Web camera images, while the network flow estimation module features a traffic flow estimation algorithm at every single link, based on measurements at a limited number of links with the cameras. Using real Web cameras deployed in Nairobi, Kenya, we assessed the accuracy of our approach. To the best of the authors' knowledge, this is the first practical framework for monitoring an entire city's traffic without special expensive infrastructure and time-consuming data calibration.