Facial Recognition for Attendance

Abstract

In computer vision, facial recognition has become a key technique with applications in various domains, including customised user experiences, human-computer interaction, and security. To build a facial recognition system, this project makes use of the robust OpenCV C++library. Constructing a reliable and effective system that can identify and detect faces from streaming videos and photos. The research commences by investigating the basic principles of facial recognition, and thoroughly examining preprocessing methods such as face detection, picture normalization, and feature extraction. OpenCV's extensive toolkit makes these tasks easier and guarantees quickness and accuracy in the recognition process. Robust face detection is emphasized as a prerequisite for further recognition phases in the project. The project implements facial feature representation and extraction using the Eigenfaces algorithm. Eigenfaces are a useful tool for efficient face identification because they offer a compact and discriminative representation of facial traits. Eigenface decomposition and subsequent classification are carried out by the C++ implementation by utilising matrix manipulation features of OpenCV. The study investigates additional strategies like local binary pattern (LBP) and histogram of oriented gradients (HOG) to improve the accuracy of the recognition system. By using these techniques, the system is better able to handle changes in lighting, position, and facial expressions because of the increased feature representation.