Intelligent Driver Monitoring System for Safe Driving

Abstract

More than 80 percent of collisions will result from the driver's negligence brought on by drunkenness and drowsiness. Numerous monitoring systems based on machine learning are being included in some cars. All these innovations are made in new automobiles as technology advances at a supersonic rate but leave a lot of empty spaces in each breakthrough. However, no single mechanism is now in place to assess the driver's identification, level of intoxication, and drowsiness. So this paper proposes a feasible system for monitoring real-time driver's wellness: “Intelligent driver monitoring System for safe driving” which gives suitable warnings and notifications to the user and the authority if necessary. This system includes authentication of the user followed by Drunkenness detection and Drowsiness detection. The user can interact with the Graphical user interface and the camera provided in the automobile frequently monitors the driver using Machine learning and neural network techniques, this checking method is implemented. This multi-model system uses many layers in Convolution neural network model and 92 percent accuracy is shown throughout the testing phase so that this system can replace any existing system which is only based on single models.