Driver Drowsiness Detection using AI Techniques

The modern age technology has evolved at a high pace to make human lives at ease. The leading mode of transportation and transferring of goods from one place to another in 1900's was bicycles, then came along motorbikes which then evolved to 4-wheeled vehicles. In today's era, it's difficult to imagine a life without a vehicle. Although it has a lot of advantages, there are definitely some life-threatening cons. According to statistics, every year around 20% of the accidents are fatigue and distraction based. Driving when one is tired makes the person prone to get involved in an accident, putting theirs as well as others life in danger on the road. Hence, some moderation is required for safekeeping of people on the road. The objective of this paper is to tackle the grave issue of drowsiness and fatigue by detecting the mouth and the eye region of the driver along with building a ‘Driver Drowsiness Detector’ prototype. The prototype was devised in various phases and sub steps, where python language has been utilised for recognising the face and eye region in a still image along with capturing the real time footage of the person sitting in front of the camera. Yawn Detection and EAR (Eye Aspect Ratio) were integrated and implemented in the prototype for detecting fatigue; which was further merged with CNN model to improve the accuracy of the drowsiness detection model.