Live Class Monitoring Using Machine Learning

Abstract

Facial recognition is a propitious tool for consumers of social infrastructure electronics. Digital cameras and mobile phones are being successfully used in recent years for digital processing of facial imagery. In this research paper, we have presented a Live class monitoring system using a face recognition algorithm which is capable of (a) processing image of a student in real-time camera-input environment using OpenCV, (b) saving the details in the backend using SQLite, and (c) keep updating the presence and absence of the student on the basis of trained dataset in Live database using Firebase. A person with the authentication, i.e., person with a dedicated QR code, can also monitor the presence and absence of the students in a class through the mobile application developed using App Inventor which is Google’s open-source platform. The dedicated mobile application will show the image of the student if he/she is present in the class and will gray out the image of the student if the opposite happens, i.e., if the student is not present for a specified time period. OpenCV library is used to import some important functions for recognition. Haar classifier and local binary pattern histogram has been used to detect the face. To store the details, SQLite3 has been used. The algorithm is applied real-time condition. The language used is Python3.