A Machine Intelligence Model to Detect Drowsiness for Preventing Road Accidents

Road accidents are one of the major issues to be addressed in recent days. Drowsiness while driving the vehicles is one of the major causes among them. Statistics state that drowsiness alert system is required to avoid this kind of problems in daily life. Few earlier inventions based on the movement of the vehicle and steering wheel angle which is not sufficient to prevent the damage. This paper proposes drowsiness identification system that identifies driver's drowsiness by capturing the image frames of the drivers face from the video stream depending on eye blinks and blinking interval. In this research work, D-CNN based model is suggested to detect the drowsiness and validate the system based on different circumstances like driver with and without glasses. Proposed D-CNN outperforms than other CNN models and able to detect drowsiness with 95% accuracy.