According to the National Highway Traffic Safety Administration, in 2017 drowsy driving resulted in 50,000 injuries across 91,000 police-reported accidents, as well as almost 800 deaths. Through the application of visual and radar sensors combined with machine learning, this research developed a drowsy driver detection system aimed to prevent potentially fatal accidents. The working prototype of Advanced Driver Assistance Systems can be installed in present-day vehicles to detect drowsy drivers with over 95% accuracy. It integrates two types of visual surveillance to examine the driver for signs of drowsiness. A camera is used to monitor the driver’s eyes, mouth and head movement in order to recognize when a discrepancy occurs in the driver's eye blinking pattern, yawning incidence, and/or head drop, thereby signaling that the driver may be experiencing fatigue or drowsiness. The micro-Doppler sensor in the system allows the driver's head movement to be captured at all times. Through data fusion and deep learning, the system quickly analyzes and classifies a driver's behavior under various conditions in real-time monitoring. This research could be implemented to reduce drowsy driving, thereby, making the roads safer for everyone and ultimately saving lives.
Transportation Engineering, Transportation Technology
Digital Object Identifier
Mineta Transportation Institute URL
Intelligent Transportation Systems (ITS), Drowsy Driver Detection, Applied Machine Learning, Accident Prevention, Advanced Driver Assistance systems (ADAS)
Other Computer Engineering | Transportation
Hovannes Kulhandjian. "Detecting Driver Drowsiness with Multi-Sensor Data Fusion Combined with Machine Learning" Mineta Transportation Institute Publications (2021). https://doi.org/10.31979/mti.2021.2015