Driver Monitoring System Using Machine Learning

As technology improves and becomes more novel, means of transportation are becoming more sophisticated. There are some rules that all motorists must follow, regardless of social status. The proposed system aims to reduce the number of accidents caused by driver drowsiness and fatigue and increase transport safety. This has become a common cause of accidents in recent years. Multiple facial and body gestures, such as tired eyes and yawning, are considered signs of driver drowsiness and fatigue. EAR (Eye Aspect Ratio) calculates the ratio of distance between horizontal and vertical eye marks required to detect drowsiness. It uses machine learning algorithms to identify facial features and alerts the driver with a buzzer when drowsiness is detected. We used Convolutional Neural Networks, a class in OpenCV and deep learning. We also use image processing, which uses computer algorithms to perform image processing on digital images. It is a camera-based technology that monitors driver attention. A convolutional neural network (CNN) is used to classify the state of the eyes and mouth. In machine vision- based driver fatigue detection, blink frequency and yawning are key indicators for evaluating driver fatigue. This project was undertaken to provide data and a different perspective on current issues.