Face Recognition Attendance System (FRAS) using CNN and Haar Classifier

shahzad nasim, Syed Muhammad rafi, Mohsin Khan, Sheikh Muhammad Munaf, Abdul Kabeer Kazi, Muhammad Ashraf


The soul purpose of this article to make an automated attendance system using face recognition techniques. It is a monitoring and controlling system for educational institutions to upgrade their usual manual attendance system. This must be more efficient and reliable as compared to previous studies. This system also consider the COVID19 scenario where human touch is involve during attendance, this system is more reliable in this situation. Because due to automatic face recognition system, human touch is not involed in any case and this will avoid some major symptoms of COVID19. Convolutional Neural Network is the major theme of this research. CNN is a Deep learning technique that detect human face in very efficient and reliable way. Firstly the database will be created and all the basic information (student enrollment, bio-data, student image etc) of all students of entered into the database. Class will be created of every batch and their course and teacher information is also added. Now Haar cascade Classifier is used to extract the features of stored images for identifying the landmarks and to complete the face detection process. Then Convolutional Neural Network (CNN) is applied on these features and extract more features. In the end face is recognized after the training process of images. For COVID19 scenario face mask images are used for face detection and face recognition. After recognition student attendance is recorded into the database. So camera is required in every class that  capture a single shot of all class students, then system will compare all these faces against the database stored images and record the attendance of every student of that particular class. In the end results will also compare against the face techniques of face recognition system

Full Text:



  • There are currently no refbacks.