Title: Driving behaviour recognition based on orientation and position deviations

Authors: Wei Sun; Xiaorui Zhang; Xu Zhang; Xiaozheng He; Guoce Zhang

Addresses: School of Automation, Nanjing University of Information Science & Technology, Nanjing, 210044, China; Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing, 210044, China ' Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing, 210044, China; Jiangsu Engineering Center of Network Monitoring, Nanjing University of Information Science & Technology Nanjing, 210044, China ' School of Automation, Nanjing University of Information Science and Technology, Nanjing, 210044, China ' Department of Civil and Environmental Engineering, Rensselaer Polytechnic Institute, Troy, NY, 12180, USA ' School of Automation, Nanjing University of Information Science and Technology, Nanjing, 210044, China

Abstract: This paper proposes a driving behaviour recognition method, which applies vehicle orientation and position deviations to warn the driver against possible dangers. We integrate a gradient reinforcement method based on the linear discriminate analysis (LDA) to reinforce lane edges. An improved Canny operator based on adaptive threshold segmentation is exploited to extract the lane edges reliably. Based on an improved Hough transform algorithm, the reinforced lane edges help the detection of polar angle and polar radius of lanes that are used to calculate the vanishing point position. After that, the proposed method predicts current-frame lane parameters based on the previous-frame parameters through using the Kalman filter. Combining deviation angle and deviation distance, the proposed method categorises vehicle lane-keeping behaviour into three states: normal, left deviation, and right deviation. Experimental results of a variety of travelling scenes show that the proposed method outperforms other existing methods in precision.

Keywords: lane detection; vanishing point; Hough transform; Kalman filter; slope angle; offset distance; fatigue driving.

DOI: 10.1504/IJSNET.2019.100219

International Journal of Sensor Networks, 2019 Vol.30 No.3, pp.161 - 171

Received: 26 Feb 2019
Accepted: 09 Mar 2019
Published online: 18 Jun 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article