Driver Hypo-vigilance Detection Based on Eyelid Behavior

Abstract

Driver face monitoring system is a real-time system that can detect driver fatigue and driver distraction using machine vision approaches. In this paper, a new algorithm is proposed for driver hypo-vigilance detection based on eye-region processing and without explicit eye detection stage. In this method, horizontal projection of top half-segment of facial image is used to extract symptoms of fatigue and distraction. Percentage of eye closure (PERCLOS) and eyelid distance changes during time are used for fatigue detection; and eye closure rate is used for distraction detection. The novelty of our method is in adaptive feature extraction using spatio-temporal processing without explicit eye detection. Processing rate of proposed method is more than 5 frames per second.