AI-Driven Real-Time Driver Drowsiness Detection for Safer and Sustainable Transportation

Driver restlessness, a major contributor to traffic accidents, must be detected to enhance driving safety. Previous methods relied on driver self-evaluations or occasional behavior-based assessments, both of which suffer from accuracy and consistency issues. Self-assessments are often unreliable due to cognitive bias, while infrequent monitoring may fail to detect drowsiness in real-time. This study introduces an innovative approach utilizing MediaPipe and OpenCV for real-time drowsiness detection. OpenCV facilitates image processing and feature extraction, while MediaPipe, a Google-developed framework for multimodal machine learning, tracks facial landmarks and analyzes eye movements. By leveraging MediaPipe's advanced facial recognition, the system monitors eye closures and head positions that signal fatigue. OpenCV further improves the detection algorithm's accuracy and responsiveness through precise image processing. Experimental results validate the system's effectiveness in identifying drowsy states, highlighting its potential for integration into in-vehicle safety technologies.