Lane detection and tracking by monocular vision system in road vehicle

Abstract

In this paper, we address the problem of lane detection and lane tracking system by a single camera CCD, installed on-broad of a vehicle for driving assistance or autonomous vehicle control in road environments. A lane detection algorithm based on Support Vector Machines (SVM) classifier for marked or unmarked roads is proposed and Catmull Rom splines based lane model combined with a simple Kalman filter for the tracking algorithm. The results of the experiment are evaluated on a number of road scenes images, and show the performance of our proposed approach.