Online Attention Detection And Judgement Based On Facial Features

In recent years,under the background of COVID-19 and "Internet + education",online learning has become an important auxiliary form of education for the educated group,and has gradually become systematic and normalized from the original freshness.Students’ attention to online learning and their investment in learning all affect the quality of online teaching.Due to the lack of "classroom presence",the lack of effective supervision and guidance,and the influence of factors such as autonomous learning ability and their own learning will,attention to teaching content has declined,which affects the quality and effect of their learning.Therefore,the study of students’ attention to online learning has important practical significance for the development of online education and teaching.The online attention detection and judgment method based on facial features combines the characteristics of eye fatigue and eye attention state to study the online attention.Through face recognition,eye location,face correction,obtaining the center position of iris and pupil,extracting and analyzing the fatigue characteristic parameters and gaze direction in real time,and completing the online attention judgment.The main contents of this paper are as follows:（1）The face detection method based on Histogram of Oriented Gradient by Dlib library is used to detect the face in video image.In order to reduce the interference of the left and right deflection posture of the head on the judgment of the gaze direction,the tilt angle between the straight line and the horizontal line where the two outer corners of the left and right eyes are located is calculated to analyze whether the face faces the screen and correct the face image that does not face the screen.（2）An improved pupil center location method is proposed.The eye feature points are repositioned in the corrected face image.The eye image is obtained by using the eye image mask.The contour of the binary image of iris and pupil area is obtained through bilateral filtering and gray image corrosion,and then the position of pupil center is obtained.Through comparative test,it is verified that the improved pupil center positioning method has good positioning effect.（3）Establish the evaluation standard of online attention,and obtain the characteristic parameters of eye fatigue for fatigue detection according to the characteristics of eyes when people are tired;According to the position of the pupil center in the whole eye image,the direction of gaze is judged,so as to analyze the eye attention state.Through the fusion analysis of eye fatigue state and attention direction,judge the online attention.The experimental results show that the pupil center positions can be well located without any occlusion of the face,the eye attention status can be analyzed,fatigue can be judged,and the online attention of students can be effectively judged by online attention detection and judgement based on facial features.