基于视觉的驾驶人疲劳及注意力监测方法  

A Surveillance Method for Driver's Fatigue and Attention Based on Machine Vision

作　　者：郭克友[1] 张春雨 

机构地区：[1]北京工商大学机械自动化学院,北京100037 [2]国家智能交通系统工程技术研究中心,北京100088

出　　处：《公路交通科技》2010年第5期104-109,共6页Journal of Highway and Transportation Research and Development

基　　金：国家(八六三计划)高技术研究发展计划资助项目(2006AA11Z213);北京工商大学青年教师科研启动基金资助项目

摘　　要：对驾驶人疲劳状态监测技术进行了深入研究,分析了在实际应用中存在的问题,在此基础上提出基于灰度方差定位驾驶人面部,利用投影曲线极点位置分割面部器官独立区域,对每一个独立区域进行OSTU计算和轮廓提取,最终获得眼睛的轮廓状态,利用PERCLOS判断驾驶人的精神状态。利用驾驶人面部器官定位的结果,对人体头部旋转运动模型进行分析,提出了计算驾驶人面部旋转角度的计算方法。最后给出判断驾驶人的注意力是否分散的判断策略。试验结果表明,该算法实时性好,具有较好的鲁棒性,可以为驾驶人疲劳状态及注意力状态的判断提供较为准确的依据。The monitoring technology for fatigue state of driver and the problem in the process of actual application were studied. On this basis, a new algorithm based on gray variance was presented to locate the face of driver. This algorithm uses the location of the extreme point in the projective curve to divide the facial organs into separate areas, dealt with the areas based on OSTU algorithm and extract the organs＇ contours, and obtain the location of the eye＇ s contours. The mental state of the driver was determined by PERCLOS. Then, based on the result of the location of the facial organs of the driver, the rotary motion of the head was analysed, and an algorithm for calculating the facial rotary angle of the head was proposed. Finally, a strategy for judging whether the driver＇ s attention is disturbed or not was obtained. The experiment result confirms that this algorithm has good real-time performance, greater robustness and can provide a more accurate basis for analyzing the driver＇ s fatigue state and attention state.

关 键 词：智能运输系统 疲劳及注意力监测方法 机器视觉 驾驶人 

分 类 号：U491[交通运输工程—交通运输规划与管理]