Vision and odometry based autonomous vehicle lane changing

Abstract

Autonomous driving is an old desire of mankind. The main purpose of this article is to present an autonomous lane changing method, that enables fully automated lane changing if given safety criteria are met. Environment detection is executed by multiple sensors such as radar, LIDAR and a smart-camera. Intrinsic sensors of the vehicle are utilized as well to ensure a smooth transition between lanes. The test results obtained on a closed test track prove the usefulness of the presented methodology. The resulting algorithm and experiences collected during implementation could be useful for integration into a larger system. Keywords: Autonomous vehicle, Lane change, Vision, Odometry