JIETA, Vol.3 no.1, pp.140-144, 2020
Vision-based Drowsiness Detection and Alarm System
Fernando E. Quiroz, Jr. 1,*), Renz Joven S. Madeja 1), Christian Rick A. Jaro 1)
1) College of IICT, Biliran Province State University, Philippines
Abstract: In this paper, a vision-based system is proposed to monitor and detect drowsiness of the driver in real-time. The system uses the histogram of oriented gradient-based face detector alongside with a facial landmark detector, which is use to extract eye areas. The next phase computes the eye aspect ratio using the Euclidean distance formula. Using a threshold value of the eye ratio, the system can now identify if a driver is drowsy. For evaluation and experiment, the system was tested by fifty drivers of different vehicle types including car, bus, delivery truck, and van.
Keyword: computer vision; drowsiness detection; road safety