Computer Vision on Embedded Sensors for Traffic Flow Monitoring

Abstract

Capillary monitoring of traffic in urban environment is key to a more sustainable mobility in smart cities. In this context, the use of low cost technologies is mandatory to avoid scalability issues that would prevent the adoption of monitoring solutions at the full city scale. In this paper, we introduce a low power and low cost sensor equipped with embedded vision logics that can be used for building Smart Camera Networks (SCN) for applications in Intelligent Transportation System (ITS), in particular, we describe an ad hoc computer vision algorithm for estimation of traffic flow and discuss the findings obtained through an actual field test.