Optimization Algorithms in Biometric Recognition Systems: A Systematic Review and Comparative Analysis

Abstract

Biometric recognition systems have become an integral component of modern authentication due to their ability to provide secure and reliable identity verification across applications including banking, healthcare, border control, forensic investigations, and national identity management. As biometric systems increasingly rely on high-dimensional feature representations generated by conventional feature extraction methods and deep learning models, efficient optimization techniques have become essential for improving recognition performance while reducing computational cost. This study presents a systematic review and comparative analysis of optimization algorithms applied to biometric recognition systems published between 2020 and 2026. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines, relevant studies were retrieved from Scopus, Web of Science, IEEE Xplore, ScienceDirect, SpringerLink, ACM Digital Library, Wiley Online Library, and Google Scholar. The selected studies were evaluated according to optimization technique, biometric modality, recognition accuracy, feature reduction capability, convergence speed, computational complexity, robustness, and scalability. The review shows that optimization algorithms substantially enhance biometric recognition by improving feature selection, reducing feature dimensionality, optimizing classifier performance, and facilitating multimodal biometric fusion. Comparative analysis indicates that recent bio-inspired algorithms, particularly the Whale Optimization Algorithm (WOA) and Grey Wolf Optimizer (GWO), consistently achieve superior performance in terms of recognition accuracy, convergence behavior, and scalability across multiple biometric modalities. The review further identifies emerging research directions involving optimization-assisted deep learning, hybrid optimization frameworks, explainable artificial intelligence, and multimodal biometric systems, while highlighting key challenges and opportunities for future research.