Video-based Driving Fatigue Detection

Fatigue was mainly due to overload or extended time of physical or mental work caused by a complex physiological phenomenon. It is a normal human physiological activities of the law, but also the human body itself can not resist the state. Monotonous highway road surface conditions and the high traffic speed, this high-intensity, driver's activities can easily cause fatigue and lead to traffic accidents, which endangering the lives of the people. Although fatigue is a normal physiological phenomenon, but this special group of drivers may cause serious consequences, and even life-threatening. Therefore, the development of real-time monitoring of driving behavior and driver fatigue warning systems are essential to reduce casualties and economic losses. Therefore, the study of fatigue on driving, not only has important theoretical significance, but also has a significant social value and economic value. Developed countries start driving fatigue research early, and they has made a large number of research results. Based on the analysis of driver fatigue detection method at home and abroad, driving fatigue detection was carried out on the basis of video images.Based on the video collected on a simulated driving cab, a new algorithm of eye localization and eye state recognition in driving fatigue detection was proposed. This algorithm used the Gabor filter to smooth the image and pick up the terrain feature of the images. Then support vector machine (SVM) was used to validate the eye area. Finally, SVM was used to recognize the eye state. The robustness of the algorithm was demonstrated under various image conditions, different eye states and facial appearances. The SVM classifier can determine the eye state qualitatively and quantitatively. The result of the algorithm provides necessary data for the future analysis of driving fatigue. Based on the video images to determine the status of the human eye can found the differences of awake and fatigue. Blinking frequency is the same, the blink time is growth.This research is sponsored by National Science and Nature Foundation of China (Porject No. 60422201).