基于眼睛状态识别的疲劳驾驶检测

由于光照变化、头部姿态等因素的影响,现有的疲劳驾驶检测精度仍有待提高。针对该问题,提出一种基于迁移学习的眼睛状态识别网络(Gabor and LBP-convolutional neural networks,GL-CNN),该网络是由Gabor特征和LBP特征通过迁移学习加入卷积神经网络(convolutional neural networks, CNN)调制组成的。首先用多任务级联卷积神经网络(multi-task CNN,MTCNN)检测驾驶员的人脸和双眼,然后经过眼睛筛选机制获取待检测的单只眼睛,通过GL-CNN识别眼睛的睁闭状态,最后根据PERCLOSE准则判断驾驶员的疲劳状态。实验结果表明,该算法具有较高的准确率,可以检测多种姿态眼睛的状态,同时满足实时性的要求。

Fatigue Driving Detection Based on Eye State Recognition

Due to factors such as illumination variation and head posture, the accuracy of existing fatigue driving eye state detection needs to be improved. Aiming at this problem, an eye state recognition network GL-CNN (Gabor and LBP-Convolutional Neral Networks) based on transfer learning was proposed. The network was composed of Gabor features and LBP features added to CNN (Convolutional Neral Networks) modulation through transfer learning. First, the driver''s face and eyes were detected by the multi-task cascade convolutional neural network MTCNN (Multi-Task Convolutional Neral Networks). Then, through the eye screening mechanism, a single eye to be detected was acquired, and the opening and closing state of the driver''s eyes was identified by GL-CNN. Finally, the driver''s fatigue state was judged according to the PERCLOSE criterion. The experimental results show that the algorithm has higher accuracy and can detect the state of multiple posture eyes, and meet the requirements of real-time.