Real-Time Android Application for Traffic Density Estimation

This paper deals with automatic traffic density estimation using new technologies that are cost-effective, quick to deploy, and easy to adapt. Therefore, a real-time video processing-based android application (app) is developed. It can be used online by exploiting the smartphone's camera placed at the roadside or offline and by using pre-recorded vehicles flow videos and an Android emulator. The system autonomously computes the vehicle flow density in real time and saves all the relevant information in the smartphone memory. The developed app is validated under real conditions in Sherbrooke and Montreal cities, by considering different meteorological conditions and varying intrinsic and extrinsic camera parameters. Obtained results allow being optimistic about the effectiveness and the applicability of the proposed app.