Face Recognition Based Attendance System using Haar Cascade and Local Binary Pattern Histogram Algorithm (KEY IJP************164)

• S. Md. Riyaz Naik,P. Jayanth,N. Sai Murali Krishna Yadav,S. Kareem Basha,V. Sumanth Reddy,N. Abdul Vasi

Abstract

Abstract: An attendance system is crucial for monitoring student presence in classes, with various methods available such as biometric, RFID card, face recognition, and traditional paper-based systems. Among these, face recognition stands out for its security and efficiency. This research focuses on enhancing the face recognition attendance system's accuracy by minimizing false positives through a confidence threshold based on the Euclidean distance metric. Because it can handle color changes well, the Local Binary Pattern Histogram (LBPH) algorithm works better than other distance-based methods like Eigenfaces and Fisherfaces. Haar cascades are used for face identification because they are reliable, and LBPH is used for classification. It works 77% of the time to recognize kids and 28% of the time to give the wrong answer. Notably, it can spot students even if they have differences, like glasses or facial hair. Face recognition for unknown people is about 60%, with 14% and 30% false positives with and without the barrier, respectively.