Real Time Vision Based Road Lane Detection and Tracking

Taking into account these various remarks, our method is composed of four modules: In this article, we describe a new method deA recognition phase allows to accurately detersigned to detect and track the road lane of a vehicle mine the position of the road sides in the image. equipped with a monocular monochromatic camera. This phase is based on the exploitation of a staThe goal of this work is to assist the conductor in tistical model of the lane. the framework of a trajectory supervision and conThe method is applied both on marked and nottrol. The work described concerns the detection of marked roads images. This step is described the roadsides in the image based on the exploitation more precisely in [I]. of a statistical model of the lanes, the localisation of the vehicle on its lane and the management of the number of lanes. The method presented here has been implemented on our experimental vehicle VELAC. Results obtained show the robustness and the precision of the proposed approa...

• Type: Conference
• Year: 2000
• Where: MVA
• Authors: Romuald Aufrère, Roland Chapuis, Frederic Chausse
• Keywords: Method DeA Recognition, Monocular Monochromatic Camera, MVA 2000, MVA 2007, Work Described Concerns