Design And Implementation Of Teaching Feedback System Based On Sitting Posture Image Recognition

With the reform of the teaching model,many new teaching modes have begun to emerge in China,but traditional classroom teaching is still the most important teaching method.In the classroom teaching,the teacher judges whether the students are devoted to the classroom by observing the students’ postures,expressions,facial expressions and other information.The information belongs to the teaching feedback information.Teaching feedback is an important guarantee for optimizing teaching results,and it has a positive effect on both students’ learning and teachers’ teaching.At present,this kind of instant teaching feedback information can only be observed by the naked eye of the teacher,it is very difficult to rely on manpower for complete collection.Aiming at the current situation of teaching feedback information collection,a teaching feedback system based on sitting posture image recognition is proposed.The image recognition technology is used to automatically count the number of students in different sitting postures in the classroom.The statistics of student sitting posture are used as feedback information,and the collected information is analyzed and displayed.The purpose of this study is to let the computer do the collection of teaching feedback information instead of teachers,and to show the quality indicators of the course to the teaching managers for reference.The system uses image recognition technology to complete the automatic collection of teaching feedback information,statistical analysis of the collected information,calculate the student’s concentration,and visually display the results in the form of a chart.This paper mainly expounds the design and implementation of the system.The research tasks include the analysis of the main processes of the system,the architecture design of the system,the design of each module of the system,the construction of the neural network model,and the key codes realized by each module.The artificial neural network used in the system is Faster R-CNN,which is used to recognize the sitting posture of students.The system detects the sitting posture of the student in the image captured by the camera,and recognizes and classifies each sitting posture,and obtains the number of students in each sitting posture in a frame of the image.This process begins every class and occurs once per second until the end of the class.The system collects the sitting position statistics of the whole class,from which the results indicating the degree of student concentration can be calculated.Finally,the statistical results can be displayed in the form of a chart on the browser.This study describes in detail the design and implementation process of the teaching feedback system.Firstly,the system is analyzed for requirements.According to the analysis results,the overall architecture design and detailed module design are carried out.Then there is the implementation of the various functional modules of the system.Among them,the focus is on the part of the sitting image recognition,including the collection of samples,the processing of samples,the training of models and the testing of models.The implemented system achieved the desired goal,was able to recognize the student’s sitting posture image with a set frequency,and had a available accuracy.The concentration score given by the system is close to the actual situation,and it can provide reference for the evaluation of teaching quality.