IoT and Computer Vision Based Driver Safety Monitoring System with Risk Prediction

A system to monitor driver safety was developed to alarm the driver when he becomes drowsy and also to centrally update driver condition in a database. To measure the concentration level of the driver we averaged few factors such as drowsiness, his face in frame. As most of the methods lack accuracy we used robust facial landmark detectors to detect face in frame and to also detect eyes. By few algorithms, driver drowsiness and drivers face position in frame was detected to alarm the driver locally. At the same time data was uploaded to a database for further inspection. We averaged the data to measure their concentration level and finally made a prediction by using their previous data. This system can be used to notify the driver and take proper actions to prevent the accident before happening.