基于PERCLOS的机动车驾驶员驾驶疲劳的识别算法

阐述了PERCLOS测评驾驶疲劳的机理，对测试驾驶疲劳的几种方法的测试精度进行了比较，认为P80是最好的。应用二维高斯模型、灰度直方图、灰度模式匹配等图像分析和识别手段定位和追踪眼睛睁开、闭合的变化过程，统计出眼睛闭合时间。利用概率和数理统计方法给陆了一种行之有效的、基于PERCLOS的机动车驾驶员疲劳程度测的新算法。

PERCLOS-Based Recognition Algorithms of Motor Driver Fatigue

A brief review of detecting and evaluating technique of motor driver fatigue is provided. According to the comparison between several commonly used detecting and evaluating techniques of motor driver fatigue, the P80 of PERCLOS (the proportion of time the eyes were closed at least 80 percent) is most significantly correlated with driver fatigue. A convincing case has been made that slow eyelid droop PERCLOS has the best potential to detect fatigue. A feasible method based on the mathematical model and symmetry analysis is presented to detect and locate the driver's eye in an image and the gray scale mode matching technique is used to determine the driver fatigue degree. An example for detecting the eyelid closure over the pupil over time is given. The combinations of mathematical model and symmetry analysis increases the robustness of the performance while the target is deformed. The scheme is suitable for human face location in intelligent human machine interface and provides a groundwork to practical application of face recognition techniques. The PERCLOS based recognition algorithms will be considered as a most effective method used to accomplish real time measure of alertness for in vehicle, drowsiness detection systems.