Optimizing Face Recognition Accuracy through Eigenfaces and K-Nearest Neighbour: An Empirical Study

Abstract

Face recognition has emerged as a cornerstone of computer vision research, spurred by its expansive utility in security systems, surveillance, and human-computer interaction. This study develops a robust face recognition framework that integrates the Eigenfaces methodology for sophisticated feature extraction with the K-Nearest Neighbour (KNN) algorithm for classification. By leveraging Principal Component Analysis (PCA), Eigenfaces are generated to significantly condense the dimensionality of facial datasets while preserving the most salient discriminative features. These optimized feature vectors are subsequently processed via a KNN classifier, utilizing a range of K values to perform a thorough evaluation of the system's operational performance. The empirical findings reveal that the proposed methodology yields high recognition accuracy alongside stable computational overhead, suggesting that the selection of an optimal K value is essential for balancing predictive precision and processing efficiency.