Eye-blink rate detection for fatigue determination

This paper focuses on detection of eye, calculating the eye-blink rate of a person and then determining his level of attentiveness. It is established by researchers that the level of attentiveness of a normal person is greatly expressed by his eye-blinking rate. If a person is feeling fatigued or drowsy, his eye blinking pattern changes. In this paper BioID dataset has been employed for eye detection and result were 97% positive detection. Using the detected eyes, we determined the eye-blink rate which if varied from the average threshold level determined that the person is fatigued or low on attention. The result for eye-blink detection is 87%. This methodology when employed in vehicles will help greatly in reducing the amount of fatigue related accidents by alarming the driver so that he can take preventive measures.