摘要
随着升学竞争越来越大,学生负担也越来越重,为了提高教学质量和升学率,需要对课堂上学生状态进行判断,找出处于疲劳状态的学生。然而,这些方法不能同时保证检测的准确性和速度。因此,文中提出了一种基于多特征融合的疲劳检测方法。首先,利用MTCNN改进了基于MedianFlow的人脸跟踪算法。然后提出了一种新的基于CNN的人脸关键点检测模型,利用人脸关键点检测的结果对眼睛进行定位。最后,通过融合闭眼时间、眨眼频率和头部位置等信息来实现学生疲劳的检测。实验结果表明,文中的疲劳检测方法在速度和精度上都有很好的效果。
With the increasing competition for higher education,the burden of students is getting heavier.In order to improve the quality of teaching and the rate of higher education,it is necessary to judge the status of students in the classroom and find out students who are in a state of fatigue.However,these methods cannot guarantee both accuracy and speed of detection.Therefore,this paper proposes a fatigue detection method based on multi-feature fusion.Firstly,the face tracking algorithm based on MedianFlow is improved by using MTCNN.Then a new CNN-based facial key points detection model is proposed,which uses the results of facial key points detection to locate the eyes.Finally,the student’s fatigue is detected by fusing information such as closed eye time,blink frequency and head position.The experimental results show that the fatigue detection method proposed in this paper has a good effect on speed and accuracy.
作者
李芙蓉
LI Fu-rong(Shaanxi Vocational&Technical College,Xi’an 710038,China)
出处
《信息技术》
2020年第6期108-113,120,共7页
Information Technology
关键词
人脸追踪
人脸关键点检测
眼睛位置
疲劳检测
face tracking
face key points detection
eye location
fatigue detection
作者简介
李芙蓉(1990-),女,硕士,研究方向为信息化建设。
