Research On Key Technology Of Non-contact Class Attendance Based On Dynamic Trajectory Analysis

Attendance management is of great significance to students,teachers and management departments.Many studies have shown that there is a significant positive correlation between students' attendance rate and their academic performance;by studying students' attendance records,we can judge their inclination and determination to the curriculum.Attendance management is one of the effective means to improve the attendance rate.With attendance,students' behavior problems can be found in time.In the current education scene,there are still some disadvantages in the attendance work based on the traditional technology:1.Fingerprint based,although the anti-counterfeiting is high,but the fingerprint sensor is easy to wear and tear.and the health and safety is low for a long time;2.Face recognition based,though highly efficient,is greatly affected by the environment;3.Radio frequency or Bluetooth based,strongly depends on the device,if the device is forgotten or lost,it will It will bring a lot of troubles;4.The traditional methods are mostly one-time detection,which can not monitor the dynamic changes of attendance and location in the course of teaching in real time,such as whether the students have early leave or location changes.Based on the in-depth investigation of the relevant literature at home and abroad,and the comparison of the advantages and disadvantages of the attendance system based on fingerprint,face recognition,bluetooth and rf technology,this paper proposes a non-contact attendance model.This model combines the advantages of face recognition and millimeter wave technology.On the one hand,it uses face recognition technology to dynamically complete identity recognition in the natural environment;on the other hand,it uses millimeter wave technology to carry out high-precision detection of students'velocity and position,so as to fully grasp the classroom dynamics while improving the efficiency of attendance.Specifically,in terms of technology,this paper uses face recognition algorithm to process the collected images and realize identity recognition;uses DBSCAN clustering algorithm and Kalman filter to process the extracted distance,velocity and angle features,and realizes human body positioning,counting,motion track and speed tracking.In the aspect of experiment,the model proposed in this paper is verified in the actual scene.The experiment focuses on the identification,location,counting,trajectory and speed tracking of human body in the scene.The experimental results show that the comprehensive use of face recognition and millimeter wave technology can dynamically and effectively complete attendance work on the basis of space-time trajectory analysis,and real-time monitor the change of students' position during the teaching process.The non-contact class attendance based on dynamic trajectory analysis realized in this study has the advantages of high accuracy,good real-time performance,user friendliness and wide application scenarios:its distance resolution is 0.05m,speed resolution is 0.08m/s;it can obtain the information of moving trajectory and location in real time;the whole attendance process does not need users to carry or touch any equipment,almost in the form of no perception by users It can be operated in strong light,dark and sheltered environment.