Publication:
Driver drowsiness detection and monitoring system

Loading...
Thumbnail Image
Date
2023-08
Authors
Lee, Jee Shen
Journal Title
Journal ISSN
Volume Title
Publisher
Research Projects
Organizational Units
Journal Issue
Abstract
Drowsiness driving has been a significant risk to road safety, leading to accidents and loss of life. To address this issue, a non-invasive, real-time Driver Drowsiness Detection and Monitoring System (DDDMS) with appropriate alert and Autonomous Emergency Brake (AEB) system are proposed. This research had also emphasized the objective of evaluating the performance of using computer vision techniques for drowsiness detection. The developed DDDMS utilizes computer vision techniques, combining Haar Cascade and Dlib shape predictor 68 face landmarks to detect signs of drowsiness such as closed eyes, yawning, and head tilting. The system operates on a Raspberry Pi equipped with a camera to capture the driver's facial images and analysed them for drowsiness indicators. When signs of drowsiness are detected, the system alerts the driver using auditory cues through a speaker and takes measures to ensure safety, such as slowing down the vehicle using Pulse Width Modulation (PWM) or applying brakes with emergency lights activated. The performance evaluation of the DDDMS demonstrates high effectiveness in drowsiness detection achieving overall detection rates and activation rates for alert and AEB systems above 80% and 85%, respectively. User feedback also indicates satisfactory experiences with the developed DDDMS. Overall, the research highlights the system's potential in enhancing driver safety and reducing the risks associated with drowsy driving, making it a promising solution for addressing driver drowsiness issues on the roads.
Description
Keywords
Citation