疲劳驾驶多源性智能监测预警方法研究

Study on multi-source intelligent monitoring and early warning method for fatigue driving

摘要:

对疲劳驾驶监测预警方法进行研究,可以避免驾驶员因疲劳驾驶产生的交通事故,减少因疲劳驾驶造成的人员伤亡和经济损失。当前的疲劳驾驶监测预警方法存在监测灵敏度低、可靠性差等问题,不能及时对疲劳驾驶的驾驶员进行报警,来避免交通事故的发生。为此,提出了疲劳驾驶多源性智能监测预警方法,首先将摄像头采集的驾驶员图像进行预处理,通过计算驾驶员图像信息的灰度值,得到驾驶员图像中像素的分布密度,为后续的监测和预警工作提供信息。其次,采用卡尔曼滤波算法对驾驶员的图像信息进行跟踪,得到驾驶员各个时间内的状态估计值,最后,通过计算驾驶员状态估计值判断驾驶员是否存在疲劳状态。实验结果表明,该方法的丢包率低、多源性高、抗干扰能力强、计算效率高。

Abstract:

The study of fatigue driving monitoring and warning method can avoid traffic accidents caused by fatigue driving, and reduce the casualties and economic losses caused by fatigue driving. The current methods of monitoring and early warning for driving fatigue have the disadvantages of low sensitivity and poor reliability. It is impossible to alarm the drivers in time to avoid the occurrence of traffic accidents. To this end, put forward the fatigue driving multi-source intelligent monitoring and early warning methods, the driver collected camera images preprocessing, gray image information by calculating the driver value, distribution density in the image pixel driver, to provide information for the follow-up monitoring and early warning work. Secondly, tracking using Calman filtering algorithm on the driver image information, each time the driver in the state estimation, finally, to determine whether the driver fatigue state by estimating the driver state value. Experimental results show that the proposed method has low packet loss rate, high diversity, strong anti-interference ability and high computational efficiency.