Related Experiment Video
Updated: Aug 23, 2025
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
4.0K
Computer vision-based approach to detect fatigue driving and face mask for edge computing device
Ashiqur Rahman1, Mamun Bin Harun Hriday1, Riasat Khan1
1Department of Electrical and Computer Engineering, North South University, Dhaka, Bangladesh.
Heliyon
|November 3, 2022
Summary
This study introduces an embedded system using computer vision and heart rate monitoring to detect driver fatigue in real-time. The system achieves high accuracy in identifying drowsiness and detecting face masks, aiming to reduce road accidents.
Area of Science:
- Computer Vision
- Embedded Systems
- Artificial Intelligence
Background:
- Road accidents cause alarming fatalities and socioeconomic losses globally, with fatigued driving identified as a major contributing factor.
- Modern automotive technologies offer potential solutions for mitigating road accident severity and frequency.
- The COVID-19 pandemic highlighted the need for integrated driver monitoring systems, including face mask detection.
Purpose of the Study:
- To develop and evaluate a comprehensive embedded system for real-time fatigue driving detection.
- To integrate computer vision and heart rate monitoring for enhanced driver state analysis.
- To incorporate face mask detection into the system for pandemic-related safety compliance.
Main Methods:
- Utilized an Nvidia Jetson Nano developer kit and Arduino Uno for system processing and control.
- Employed deep learning and computer vision techniques, including analysis of 68 facial landmarks (eye and mouth localization), for fatigue detection.
- Integrated an AD8232 heart rate module for continuous monitoring of driver's heart rate variability (BPM).
- Developed a convolutional neural network (CNN) framework for face mask detection.
Main Results:
- The system demonstrated high accuracy, achieving 97.44% in fatigue detection and 97.90% in face mask identification.
- The system proved robust in various conditions, including different driver distances, lighting, eyeglasses, and oblique projections.
- Real-time monitoring and analysis successfully identified drowsiness and generated timely alarms.
Conclusions:
- The developed embedded system effectively detects driver fatigue and face mask usage with high accuracy.
- The system's ability to perform under diverse conditions indicates its practical applicability in reducing road accidents.
- This technology presents a significant advancement in driver safety and public health monitoring within vehicles.
