Multicriteria Fatigue Detection with Rapid Facial Tracking and Individualized Thresholds

Fatigue driving is one of the main causes of traffic accidents. Detecting fatigue driving can identify the possible fatigue state of the driver and issue warnings to prevent potential dangerous situations. At present, there exist challenges for available detection methods, which contains lack of real-time, low accuracy and robustness. As is a solution to problems above, the method proposed in this paper improves comprehensive performance by raising 3 new algorithms: (1) rapid face tracking algorithm, which can locate and track the driver's facial region in real time under complex lighting and background conditions, improving the detection efficiency and accuracy; (2) multi-criteria fusion algorithm, which can comprehensively consider various physiological and behavioral features, as well as their interrelationships, and output a comprehensive fatigue score, improving the detection sensitivity and robustness; (3) individualized threshold algorithm, which can dynamically adjust the fatigue threshold according to the individual differences and habits of each driver, improving the detection adaptability and personalization. The experiment results with different subjects under various fatigue driving behaviors demonstrate that the proposed approach is more accurate, fast and robust than existing ones in terms of driver identification recognition and fatigue detection.