A method for lane recognition using active contours model in vehicular roads

In this paper we propose a method for lane detection in vehicular roads using images from cameras on board a vehicle. For this purpose we used a set of recorded sequences in road scenarios with a scaled vehicle. Using the Inverse Perspective Mapping on captured image sequences and an Active Contour Model based approach adding a new internal energy parameter related to features of road such as width, number of lanes an previous observed images in order to better approximation lane lines position even with line occlusion in images.