Fatigue Estimation through Face Monitoring and Eye Blinking
Roy Amit, Gautam Paul, Milon Kanti Raha · 2014
Face detection is the general case of face localization. It is a human-computer interface technology that determines the location of faces in arbitrary images. It detects face from input image from camera. From the face template eye pair is detected, where this system automatically locates the user‟s eye. Blinks are related to several environment states and the present report describes a simple, reliable way to measure blinks from the video stream of an eye obtained during eye tracking, where the source of the eye video is a video camera attached to a head-mounted eye tracker. Computer vision techniques are employed to determine the moments that a blink starts and ends for the purpose of calculating blinking rate. The video is first processed to show blocks of eyelid and pupil movements and is then analyzed for blink starts and ends. The moment of a blink start is reported when the eyelid starts to move quickly exceeding a predetermined threshold. The end of a blink arises when the pupil size increases by less than a separate threshold. Several different blink patterns from different stage are measured. Drowse condition is determined by the analysis of blinking rate from different situation. The system has ability to count blinking rate accurately around 94%. In this study observed that if the blinking rate is approximately 3 blink/minute then it can be considered as drowse condition.