IoT-Based Driver Drowsiness and Fatigue Detection System


  • Hazizi Satiman Universiti Tun Hussein Onn Malaysia
  • Ariffuddin Joret


Drowsiness Detection System, Haar Cascade Classifier, Eye Aspect Ratio, Mouth Aspect Ratio


Driving in drowsiness condition is one of the factors that accident rates in Malaysia increase involving car, lorry, etc. Many researchers have come out the ideas on how to detect drowsiness with different techniques. One of them is by using a behavioral-based drowsiness detection system. Therefore, in this study, behavioral-based drowsiness detection system has been used since this method its non-intrusive nature. This research has been conducted to evaluate the accuracy of the behavioral-based drowsiness detection system. In order to achieve the objective of this research, Haar cascade classifier algorithm, eye aspect ratio (EAR) algorithm and mouth aspect ratio (MAR) have been implemented to detect drowsiness and fatigue. The system can detect drowsiness if the value of EAR is frequently below a threshold value (0.23), and the system will alert the driver through a speaker. Based on MAR value, this system is able to determine the driver is yawning or not. It can be said that these algorithms are good enough to detect drowsiness and fatigue.




How to Cite

Satiman, H., & Joret, A. (2021). IoT-Based Driver Drowsiness and Fatigue Detection System. Evolution in Electrical and Electronic Engineering, 2(2), 403–412. Retrieved from