Authors: Ying-Che Kuo; Chin-Sheng Chang
Addresses: Department of Electrical Engineering, National Chin-Yi University of Technology, Zhongshan Rd., Taiping District, Taichung 41170, Taiwan ' Department of Electrical Engineering, National Chin-Yi University of Technology, Zhongshan Rd., Taiping District, Taichung 41170, Taiwan
Abstract: Based on computer vision technology, a lane departure warning system proposed in this research comprises three parts: the first part contains a lane model, which is constructed to mark the potential locations of the lanes in the images. The second part relates to lane marking detection on the basis of the constructed lane model, in which the edge of the lane marking has been detected through Hough transformation. Then the extended Kalman filter is used to reduce the range of the constructed lane model. Finally, a recursive operation finds the locations of the lane path. The third part is the alerting system. The deviation between the past driving path and the detected lane is used as a baseline; and grey theory is used to estimate the future driving path and to predict the deviation between time and distance. Warnings are then given in response to dangerous driving behaviour in order to avoid accidents.
Keywords: LDW; lane departure warning system; Hough transformation; grey theory; extended Kalman filter; driving path prediction; computer vision; vehicle safety; lane model; lane marking detection; collision avoidance; dangerous driving; driving behaviour; modelling.
International Journal of Vehicle Safety, 2014 Vol.7 No.2, pp.168 - 188
Received: 08 Feb 2013
Accepted: 10 Sep 2013
Published online: 26 Feb 2014 *