Abstract

Drowsiness in four-wheeled drivers is one of the factors that cause traffic accidents. Drowsiness can be caused due to the tiredness of the journey that is passed by the driver. Utilization of artificial intelligence can be used to detect a driver's drowsiness, one of which is by observing eye activity or condition, mouth movement, and head position while driving. By knowing all these conditions, a machine can be made that can give a warning if the driver experiences possible drowsiness. This study utilizes a camera as data input to recognize the condition of the driver through activity, eyes, mouth, and head tilt position. The system will start by detecting the rider's face, then calculating each activity of eye blinking, the number or number of mouths open due to yawning, as well as head activity through poses and tilt of the head position. Face detection is used to determine the position of the face and then detect the position of the driver's eyes, mouth, and head. Utilizing artificial intelligence with the Blazeface method which is an algorithm used to map facial positions. As well as using the EAR (Eye Aspect Ratio) method to be able to determine whether the eyes and mouth are open or closed. The results of this study obtained a face detection accuracy of 98% and the system can only detect faces at an angle of 0-15 degrees.


Keywords – Face Recognition, Mediapipe, Face Extraction, Machine Learning, Blazeface.