Eye Tracking based Real- Time Non-Interfering Driver Fatigue Detection System

摘要

Driver fatigue, cause fatal road accidents, and is a major socio-economic concern. The human body exhibits signs of fatigue, which can provide us with the information needed for fatigue detection. This study presents to develop an eye tracking based real time noninterfering driver fatigue detection system. In the proposed method Haar-like features based cascade classifiers are utilized to detect the face and eyes of the driver, skin colour based segmentation is used to calculate eye openness and Support Vector Machine (SVM) to deduce the eyes state as fatigued or non-fatigued.