Research on Real Time of Distracted Driver Detection Using Machine Learning

An accident is an incident that happens suddenly, unintentionally, and under unanticipated conditions. Studies show that drivers controlling the car while severely sleepy are at blame for almost one-fourth of all serious highway accidents, suggesting that driver tiredness causes more accidents than drunk driving. A solution can be developed for the issue of detecting and alerting of driver drowsiness by utilising the technologies of computer vision, an interdisciplinary science that deals with how by creating methods that enable computers to derive high-level understanding from digital images or videos, we can increase their intelligence. The main goal of this project is to create a non-intrusive computer vision system using OpenCV that can identify driver fatigue in a real-time video stream. If the driver appears to be drowsy, the system will inform them by playing an alarm (beep). Additionally, we suggested a system for identifying and detecting traffic signs in the Traffic Sign Detection System. It is used in a way that makes quick decisions possible for drivers. Three phases make up the suggested architecture. The first is picture pre-processing, where we size the input files for the dataset, choose the input size for learning, and resize the data for the learning phase. In the course of the recognition process, the suggested algorithm classifies the observed symbol. This is accomplished in the second step using a convolutional neural network, and the third phase entails text-to-speech translation once the second phase’s discovered sign has been delivered in text format.