A Mobile-Edge-Based Smart Driver Drowsiness Detection System

Abstract

Driver drowsiness is one of the primary causes of frequently occurring roadside accidents around the world. Some possible causes of drowsiness are medical problems, non-stop traveling on long trips and the increased comfort level in modern vehicles. Drowsiness detection systems (DDS) can help avoiding such incidents by notifying the drivers in time to stop travel. By leveraging mobile-edge technology, smart devices can be helpful to capture the driver’s facial expressions like yawning, eye closure, and head movements to detect the drowsiness level. In this presented study, we employed an artificial intelligence algorithm centered on edge computing, implemented on Android, to detect driver drowsiness. The proposed algorithm uses Google Vision image processing technology for face detection and calculation of eye aspect ratio to detect driver drowsiness. Our developed system detects drowsiness of drivers with 97.2% accuracy is more practicable, less costly, and reliable for real-time applications as the processing is faster using edge computing.