DETEKSI MICROSLEEP PADA PENGENDARA MOBIL MENGGUNAKAN HAAR CASCADE CLASSIFIER DAN CONVOLUTIONAL NEURAL NETWORK

Abstract

Accidents can occur due to several factors, one of which is human factors due to fatigue and drowsiness, resulting in microsleep while driving. Previous research utilized the Haar Cascade Classifier to detect faces in real-time from a webcam (Abidin, 2018). This research focuses on developing a microsleep detection system with several stages, including eye identification via the Haar Cascade Classifier, detection of open/closed eye conditions via a Convolutional Neural Network, and drowsiness detection based on microsleep theory. From the research results, it was found that the best light intensity was 80-450 Lux (normal to bright light conditions) with the best distance being 60 cm.
Keywords: Traffic accident, microsleep, Haar Cascade Classifier, Convolutional Neural Network