This paper develops a vision-based driver-copilot system for aiding driver to keep the vehicle within the driving lane. In the copilot system, the road image features and steering wheel angle are used as its inputs and output, respectively. An intelligent lateral controller used the table-lookup scheme is implemented. Based on the experimental results by the fixed-based simulator, this work can be concluded as following: (1) The proposed system is feasible and performs well for keeping the vehicle within the driving lane while using road image features as the control inputs. (2) In the straight road, the proposed lateral controller has better performance comparing to the human driver. (3) When the switching scheme is implemented, the driver-copilot system can successfully assist the driver to keep the vehicle in the current driving lane.
A vision-based intelligent driver-copilot for vehicle lateral control
Ein intelligenter Beifahrer zur Spurführung
1998
6 Seiten, 14 Bilder, 3 Tabellen, 10 Quellen
Conference paper
English
A Vision-based Intelligent Driver-copilot for Vehicle Lateral Control
British Library Conference Proceedings | 1998
|ENGINEERING - Vision-System: Serienreifer Copilot
Online Contents | 1999