Evaluation And Analysis Of Classroom Teaching Quality Based On Human Sitting Posture Recognition

In recent years,modern information technology has developed rapidly,and at the same time the need for secondary vocational schools to seek reforms in education and teaching has become more urgent.Introducing advanced technology into the classroom of secondary vocational schools has become a new development trend.Classroom is an important place for learners to learn,communicate and grow into talents.It has received extensive attention from families,schools,and society.The feedback information of learners in the classroom can provide a reference for adjusting teaching progress and improving teaching quality.In the classroom,the learners’ different sitting postures are an important part of the feedback information,which can reflect the quality of classroom teaching to a certain extent.Human body posture detection and recognition is an important part of computer vision technology,which has made great progress in recent years.Human body gesture recognition based on deep learning has great advantages and is suitable for classroom scenes.This paper takes the sitting posture of secondary vocational school classroom learners as the research object,establishes the standard data set of learners’ sitting posture,studies the human posture detection and recognition technology based on deep learning,constructs the sitting posture recognition network,and applies it to the secondary vocational school classroom teaching,analyzes the situation of classroom learners’ sitting posture,and explores an evaluation system which combines learner’s score,teacher’s self-evaluation and learner’s sitting posture score.The main work is as follows:1.Taking the sitting posture of classroom learners as the research object,this paper analyzes the typical sitting posture characteristics of human body,establishes the classification standard,and puts forward the sitting posture score of learners as the reference basis for classroom teaching quality evaluation.The video data of real classroom scene is collected,and the learners’ sitting posture is extracted and labeled to form the labeled human sitting posture data set.2.Using the knowledge visualization tool Cite Space to analyze the development process of human posture recognition,on the basis of advanced human posture detection and recognition technology,combined with the classroom teaching scene of secondary vocational school,a fast and accurate sitting posture recognition network based on deep learning is proposed.Based on Open Pose,an open source project,and combined with the data set of human sitting posture in class,the network training of sitting posture recognition is carried out.3.Taking the classroom teaching in secondary vocational schools as an example,this paper collects the video data in the real classroom scene,analyzes the sitting scores of different classroom learners through the sitting recognition network,integrates the sitting scores of learners into the classroom teaching quality evaluation of secondary vocational schools,and explores the evaluation system of combining learners’ scores,teachers’ self-evaluation and learners’ sitting scores,and put forward suggestions and opinions to improve the quality of classroom teaching.