期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2016
卷号：4
期号：1
页码：918
DOI：10.15680/IJIRCCE.2016.0401072
出版社：S&S Publications
摘要：Driver in - a lertness is an important factor which is the major cause for the vehicle crashes. This study reviews the development of a drowsiness detection system using the concepts based on non - intrusive machine vision. A small security camera is implemented in the system which points towards the driver's face for monitoring the driver's eyes in order t o detect the drowsiness. In a particular case when the drowsiness is detected in the driver, an alert signal is given to the driver. This review study describes how to detect the eyes, and also how to determine if the eyes are open or closed. The informati on obtained is used for binary version of image for detection of eyes and face and the area is narrowed where the eyes are detected. After the detection of the face area the eyes can be detected by computing the horizontal averages in that region. Taking i n to consideration that eye region in the face show s greater intensity changes so the eyes are detected by finding these significant changes in the face. After the eyes are located the distance between the intensity changes in the eye area are measured to d etermine whether the eyes are open or closed . If a large distance in the intensity changes is measured then it corresponds to eye closure. If the eyes are detected to be closed for 5 consecutive frames, the n the system draws the conclusion that the driver is drowsy and issues a n alert signal . This system helps to detect the drowsiness and works under reasonable lighting conditions
