Design of fatigue and distraction driving detection system based on line of sight impact

Fatigue driving and distracted driving both belong to bad driving behaviors, but their hazard levels to safe driving are different. Through accurate detection and classification warning, not only driving safety can be ensured, but also transportation efficiency can be improved. According to the view angle, the driver's view is divided into 17 areas, which is more in line with the actual human physiological situation. The experiment was conducted on 30 drivers based on the indicators such as line of sight impact point, number of stay frames, and number of regional impact points, and the original data was analyzed by independent sample t-test. It was concluded that there was a significant difference between fatigue driving and distracted driving. By building a double objective prediction BP neural network model of driving state, two types of bad driving behaviors could be classified and detected with high accuracy. The warning system of fatigue driving and distracted driving was designed to realize the hierarchical warning of two types of bad driving behaviors.