基于多元数据融合的课堂表现智能检测系统设计

为了评估学生的上课行为,文章采用最新的头部姿态估计技术。此项技术能够捕捉当前人头部所做的各种动作与角度,例如低头、转头、仰头等。低头是大多数学生课堂上使用手机的惯用姿势,可以利用此项技术来监测其头部姿态从而进行评估。考虑到人脸检测的定位准确度问题,以及学生低头的动作不一定是使用手机,文章同时引入了无线信号侦测技术,捕获并检测当前教室里由于学生使用手机造成信号频繁收发的终端位置,与人脸检测的定位相结合,交叉定位,从而获得更精准的定位效果。

Designed on intelligent detection system for classroom perfprmance based on multivariate data fusion

In order to assess students'behavior in class,this paper used the latest head posture estimation techniques.This technology can capture the current person’s head to do a variety of movements and angles,such as bow,turn,head,etc.Since lowering the head is the dominant posture of most students in class,can use this technology to monitor their head posture for evaluation.Considering the positioning accuracy of face detection problem,as well as the student’s action is not necessarily the bowed their heads and the use of mobile phones,this paper also introduced the wireless signal detection technology,capture and detection current caused by students use mobile phones in the classroom signal frequency transceiver terminal location,combined with the positioning of the face detection,cross positioning,in order to gain more accurate positioning results.