Establishing a Face Recognition Research Environment Using Open Source Software
Phillip Wilson, John D. Fernandez · 2024
This paper discusses establishing a research environment for face detection and face recognition.This research environment provides real time facial detection and feature extraction from digital camera images.This environment allows research to be performed on various aspects of facial biometrics, including facial recognition, face tracking, and emotion-related feature extraction.The environment utilizes Intel's Open Computer Vision Library (OpenCV), an open source, platform independent library.This library provides the ability to analyze, extract, and modify information in real time for each frame of a digital camera or a video file.Tools are provided to create and use Haar classifier cascades to detect faces and facial features within an image or frame.OpenCV provides functions to perform principle component analysis (PCA), including covariance matrix calculations and eigenvectors calculation.The environment also utilizes the National Institute of Standards and Technology's (NIST) Facial Recognition Technology (FERET) database.This database provides hundreds of grayscale and color images of people in various lighting conditions and poses.The ultimate objective of this work is to develop an environment that can be used for multiple research initiatives related to usability and security.