人脸识别在疲劳驾驶检测中的应用研究

Application of Face Recognition in Fatigue Driving Detection

论文将卷积神经网络应用于人脸识别,对瞳孔定位算法进行改进,有效地克服了原算法计算量大的问题.根据驾驶员眼睛在不同状态下宽高比例不同的特点,实现了一种简单可行的眼睛状态判断方法,并通过PERCLOS算法对驾驶员的疲劳状态进行判定.论文采用改进Hough变换方法定位驾驶员眼睛,准确率为92%,平均响应时间为29ms,驾驶员眼睛状态的判断准确率为83.9%.设计了一种基于人脸识别的疲劳驾驶检测原型系统,实现驾驶员脸部特征检测、眼睛定位、眼睛状态判断、疲劳判定等功能.实验结果表明,系统对疲劳状态的识别率为87.5%,疲劳判断的响应时间为17ms,有较好的实际应用价值.

In the present study, convolutional neural network is applied in face recognition and improving the pupil localization algorithm, which effectively overcomes the problem of large computational complexity of the original algorithm. It is realized according to the characteristics of the driver's eyes in different states. A simple and feasible eye state judgment method is used, and the driver's fatigue state is determined by the PERCLOS algorithm. The improved Hough transform method has a positioning accuracy and average time of 92% and 29ms for the driver's eyes, respectively, and the correct rate for the driver's eye state is 83.9%. A fatigue driving detection prototype system based on face recognition was designed to achieve the functions of driver facial feature detection, eye positioning, eye state judgment and fatigue judgment. The experimental results show that the recognition rate of fatigue is 87.5%, and the response time of fatigue judgment is 17ms, which has good practical application value.