摘要
针对轨道交通列车高速行驶过程中,轨道交通列车司机因长时间工作而产生的疲劳,将影响行车安全,对机车司机作业行为以及工作条件、工作状态和监督环节重要一部分。在列车司机长时间驾驶疲劳检测其中最关键技术之一是人脸检测和人脸感兴趣区域(ROI)定位,利用高斯肤色模型对列车司机驾驶图像中的人脸感兴趣区域(ROI)进行预分割,得到列车司机的人脸感兴趣区域,提出一种面向轨道交通列车司机疲劳检测关键技术之一是人脸检测的方法,针对轨道交通列车司机疲劳检测,如人脸检测技术进行了研究,经过实验分析表明,在复杂光照下列车司机人脸方法能有效提高人脸检测率,鲁棒性较好,为后续研究精准定位列车司机的眼睛张开和闭合进行疲劳检测奠定一定基础。
In the process of high-speed running of rail transit trains,the fatigue of rail transit train drivers due to long-time work will affect the driving safety,which is an important part of locomotive drivers’operation behavior,working conditions,work⁃ing status and supervision links.In the detection of train driver’s long-term driving fatigue,one of the most key technologies is face detection and face region of interest(ROI)location.The Gaussian skin color model is used to pre segment the face region of inter⁃est(ROI)in the train driver’s driving image to obtain the train driver’s face region of interest,One of the key technologies for rail transit train driver fatigue detection is face detection.The rail transit train driver fatigue detection,such as face detection technol⁃ogy,is studied.The experimental analysis shows that the train driver face method can effectively improve the face detection rate un⁃der complex illumination and has good robustness,It lays a certain foundation for the follow-up research to accurately locate the opening and closing of train driver’s eyes for fatigue detection.
作者
江跃龙
Jiang Yuelong(Guangzhou Railway Vocational and Technical College,Guangzhou 510610)
出处
《现代计算机》
2022年第3期1-6,30,共7页
Modern Computer
基金
2019年广东省普通高校青年创新人才类项目:基于增强PERCLOS深度学习轨道交通列车司机疲劳检测方法的研究(2019GKQNCX100)
2021年市基础研究计划基础与应用基础研究项目(202102080153)。
关键词
人脸检测
疲劳检测
色彩空间
肤色分割
face detection
fatigue detection
color space
skin color segmentation