基于眼部行为的驾驶疲劳评价指标的阈值

Threshold Value of Indices of Eye States to Monitor Drive Fatigue

为将时段内眨眼次数、闭眼总持续时间和眨眼时间均值3类眼部行为指标应用于实际驾驶疲劳监测,进行了含173个样本的实验研究.将实验对象根据年龄、性别分类,以脑电指标为参照,得到各类人群在不同疲劳程度下3个眼动指标的阈值.从显著性、稳定性两个方面对3个指标进行论证分析,确定眨眼时间均值是三者中较优的指标,且验证了以年龄对人群进行划分更为合理.最后结合样本的数据,针对不同年龄人群给出3种觉醒状态下平均一次眨眼指标的阈值,通过室外车载实验验证了阈值的有效性.

In order to put eye states indices such as the frequency of blink,how long all blinks have taken,the average time of blink into practice to monitor drive fatigue,an experiment with 173 samples was set out.On the basis of classifying samples by age and sexuality,the threshold values of these three indices were calculated with the reference to EGG index.The three indices were investigated from two aspects,namely,significance of difference and stability,and then the average of blink time is regarded as the most appropriate index.Classifying people by the age is more reasonable.In the light of the data,the threshold values of this index of every crowd with different ages were figured out,and their practicality in real-time monitoring was verified in real-time monitoring through an out-of-door experiment.