Driver Fatigue Warning System Based On Computer Vision

The number of motor vehicles and drivers increase quickly in recent years. As we enjoy the use of motor vehicles, let’s be careful so that it does not have any bad effects on us. According to the statistics, as many as 1 billion fatigue-related car accidents occur every year and account for 90 per cent of all security incidents around the world. Every year fatigue driving wounds about 25 million people and kills a half million people. The road traffic accidents induced by fatigue driving will lead to severe personal casualty and property loss. According to a latest survey, Fatigued driving is an important cause of road crashes. The research on driver fatigue detection is very necessary and meaningful. The fatigue detection system works to detect and analyze whether a driver is tired or not based on real-time situation and sends out the warning if necessary. It can reduce number of fatigue-related traffic accident.Firstly, the facial images are captured by the IR CCD camera to reduce the effect of illumination. Then open eye region is located by AdaBoost algorithm with Haar-like features in this paper. Finally, this article use the PERCLOS fatigue analysis algorithm which is currently accepted as most effective method combining with the blinking frequency to judge the fatigue. The main contents of this paper include the following:1. Most often, the eye images are very small and not clear. Then a lot of eye features are ignored so that it may cause erroneous judgment. According to the structure characteristics of eyes, this paper expands Haar-like features by using AdaBoost algorithm to accurately detect the eye. Result shows that the wrong judgment due to the similitude of the eyes and eyebrow can be avoided.2. In the driver fatigue detection module, this paper comes up with a method by combining PERCLOS and blink frequency. Fatigue state could be detected by the blink frequency and the value of PERCLOS, which is the ratio of closed eye image frames and total image frames. At length, test proves that the method of this paper has good effect and high precision.