Abstract
Automatic traffic monitoring systems play an important role in the process of increasing the safety of humans and vehicles in road transportation. Vision-based approaches are promising since they allow the control of larger detection areas and the reduction of installation and maintenance costs with respect to other sensors like inductive loops which require pavement reconstruction. However, image processing techniques show robustness problems due to variable light conditions and real-time constraints required by a traffic monitoring system. An exhaustive survey of visual-based traffic monitoring systems and an overview of the principal computer vision techniques normally applied in these systems will be shown. (Author/publisher).