by Daniluk, M, Rezaei, M, Nicolescu, R and Klette, R
Abstract:
Fatigue and driver drowsiness monitoring is an important subject for designing driver assistance systems. The measurement of eye closure is a fundamental step for driver awareness detection. We propose a method which is based on eyelid detection and the measurement of the distance between the eyelids. First, the face and the eyes of the driver are localized. After extracting the eye region, the proposed algorithm detects eyelids and computes the percentage of eye closure. Experimental results are performed on the BioID database. Our comparisons show that the proposed method outperforms state-of-the-art methods.
Reference:
Eye status based on eyelid detection: A driver assistance system (Daniluk, M, Rezaei, M, Nicolescu, R and Klette, R), In Computer Vision and Gaphics, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer Verlag, volume 8671, 2014.
Bibtex Entry:
@inproceedings{daniluk2014eyesystem,
author = "Daniluk, M and Rezaei, M and Nicolescu, R and Klette, R",
booktitle = "Computer Vision and Gaphics, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
organization = "Warsaw, Poland",
pages = "171--178",
publisher = "Springer Verlag",
title = "Eye status based on eyelid detection: A driver assistance system",
volume = "8671",
year = "2014",
abstract = "Fatigue and driver drowsiness monitoring is an important subject for designing driver assistance systems. The measurement of eye closure is a fundamental step for driver awareness detection. We propose a method which is based on eyelid detection and the measurement of the distance between the eyelids. First, the face and the eyes of the driver are localized. After extracting the eye region, the proposed algorithm detects eyelids and computes the percentage of eye closure. Experimental results are performed on the BioID database. Our comparisons show that the proposed method outperforms state-of-the-art methods.",
startyear = "2014",
startmonth = "Sep",
startday = "15",
finishyear = "2014",
finishmonth = "Sep",
finishday = "17",
issn = "0302-9743",
eissn = "1611-3349",
language = "eng",
conference = "International Conference, ICCVG 2014",
}