基于计算机视觉的驾驶疲劳识别方法的研究

汽车驾驶员疲劳驾驶是引发交通事故的重要原因之一,许多国家在进行疲劳报警器的研究。在介绍和比较多种驾驶疲劳测评方法的基础上,阐述PERCLOS评价驾驶疲劳的机理,并与其他几种方法进行对比,以证明其优越性。设计一套驾驶员疲劳监控系统,应用图像差分的方法,灰度直方图、直方图均衡化等图像分析手段,定位和识别驾驶员眼睛睁开与闭合的变化过程,统计出眼睛闭合时间来确定疲劳程度,并通过利用MATLAB得以实现。给出了一种基于PERCLOS的驾驶员疲劳识别的新方法。

Detection Method for Driver Fatigue with Computerized Vision

Driving in case of drowse is a most important factor of traffic accidents. Now many countries are active in studying drowsy monitor. A lot of methods for evaluating driver fatigue are presented. According to the comparison between these commonly used detecting and evaluating techniques of motor driver fatigue, the PERCLOS is most significantly correlated with driver fatigue. A feasible driver fatigue monitoring system based on image difference, intensity histogram and intensity histogram equalization is presented to detect and locate the driver's eyes in the image and to detect the eyelid closure over the pupil over time to measure the driver fatigue degree. The method is implemented in MATLAB. A new method based on computerized vision to recognize motor driver fatigue is provided.