Citation

Teoh, Tai Shie and Em, Poh Ping and Ab Aziz, Nor Azlina (2024) Driver Drowsiness Detection Based on LiDAR-Based Road Boundary Detection. In: 2024 Multimedia University Engineering Conference (MECON), 23-25 July 2024, Cyberjaya, Malaysia.

Text 1.pdf - Published Version Restricted to Repository staff only Download (597kB)

Abstract

The road fatalities associated with microsleep have grown in recent years. To identify drowsy drivers through their driving patterns, most vehicles have adopted a front-facing camera to identify lane lines and calculate the lateral position of the vehicle. Nevertheless, the lane lines may wear off and influence the reliability of the system. Fortunately, a LiDAR sensor is capable of resolving this limitation by generating point cloud data that describes the geometrical structure of the surrounding environment regardless of weather and illumination conditions. Therefore, this paper proposes a LiDAR-based driver drowsiness detection system. Initially, the system identifies left and right road boundaries by processing the point cloud data of the LiDAR sensor. It involves ground plane segmentation, boundary point classification, and curve fitting via a quadratic polynomial. After that, the system computes the lateral position of the vehicle from the road boundaries. Based on the lateral position of the vehicle, 3 features are extracted to identify the driver's drowsiness. They are the mean lateral position, standard deviation of lateral position, and cumulative lateral position. Subsequently, these drowsiness features are passed to an artificial neural network to classify the driver's state into drowsy and non-drowsy. The experimental results showed that the proposed system was capable of achieving an accuracy of 89%. It was also discovered that drowsy drivers would have a higher standard deviation of lateral position and cumulative lateral position compared to non-drowsy drivers. Finally, the proposed system is anticipated to reduce the risks of accidents through its enhanced reliability.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Drowsiness Detection; LiDAR; Road BoundaryDetection
Subjects:	T Technology > TL Motor vehicles. Aeronautics. Astronautics
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	07 Feb 2025 00:40
Last Modified:	07 Feb 2025 00:40
URII:	http://shdl.mmu.edu.my/id/eprint/13385

Downloads

Edit (login required)