VISIONGUARD: OPTIMIZING IRIS BIOMETRIC SYSTEMS WITH MLBASED ACCURACY IMPROVEMENTS

Abstract

Iris recognition has emerged as one of the most reliable biometric authentication techniques due to its uniqueness and stability across individuals. However, traditional iris recognition methods often face challenges such as noise from occlusions, illumination variations, and feature extraction inaccuracies, which can reduce system performance. This study introduces VisionGuard, a machine learning–driven framework designed to optimize iris recognition accuracy by leveraging advanced feature extraction, dimensionality reduction, and classification techniques. By integrating convolutional neural networks (CNNs) with optimized machine learning classifiers, the proposed model enhances robustness against environmental variations and improves recognition rates across diverse datasets. Experimental evaluations demonstrate that VisionGuard achieves higher precision, recall, and overall accuracy compared to conventional approaches, thereby providing a scalable and secure solution for biometric authentication. This work highlights the potential of machine learning in advancing iris recognition systems toward real-world deployment in securitysensitive applications.