A smartphone-based driver fatigue detection using fusion of multiple real-time facial features

In this paper, a fatigue monitoring system focuses on information fusion is designed and implemented in smartphone. Eye blinking, head nod and yawning are detected as indicators of driver fatigue. We developed a mathematical model to extract the characteristic in time and frequency domain using mean-variance of key fatigue parameters. The system perform real time detection of face and eye blink using Harr-like technique and mouth detection for yawning with Canny Active Contour Method. The testing result of the system demonstrates the practical use of multiple features, particularly with our mean-variance methods, and their fusion enables a more accurate and authentic fatigue detection.