REAL TIME DRIVER STATE DETECTION

REAL TIME DRIVER STATE DETECTION

SURAJ A RAIKAR 1, SHRUTHI M T 2

1 STUDENT DEPARTMENT OF MASTER OF COMPUTER APPLICATIONS, BIET, DAVANGERE

2ASSISTANT PROFESSOR, DEPARTMENT OF MCA BIET DAVANGERE

ABSTRACT

Real-time driver state detection is a crucial aspect of modern vehicle safety and automation systems. This Python project aims to the state of a driver in real time. The project leverages learning models and image processing algorithms to analyze facial expression and head poses of the driver captured through a camera mounted inside the vehicle. The primary objective is to detect signs of drowsiness, distraction, or impairment that could compromise safe driving. By continuously monitoring these parameters, the system can issue timely alerts or take preventive actions to mitigate potential risks. The implementation involves the use of deep learning frameworks such as PyTorch for training and deploying convolutional neural networks (CNNs) for facial feature extraction and classification tasks. Additionally, real-time performance is achieved through efficient data processing techniques and optimization strategies tailored for embedded systems or onboard vehicle computers. The project's significance lies in its potential to enhance road safety by providing an automated mechanism to monitor and intervene based on the driver's real-time state, thereby reducing the incidence of accidents caused by human error. The outcomes of this research could be integrated into future autonomous vehicles or as advanced driver-assistance systems (ADAS) to augment human driving capabilities and ensure safer transportation environments.