A Bibliometric Analysis of Current Advancements in Iris Feature Extraction and Recognition

Abstract

Iris recognition has rapidly advanced due to deep learning and hybrid models that have transformed feature extraction and identification processes. However, research remains scattered across various subfields, necessitating a systematic synthesis. This bibliometric and systematic review explores recent developments in iris feature extraction and recognition by analyzing literature published over the past seven years. Using bibliometric mapping, the study identifies leading authors, journals, and collaborative networks shaping the discourse. Technological advancements, including non-segmentation deep learning models, attention-based mechanisms, cross-spectral recognition, and hybrid approaches integrating periocular features, are critically examined. Additionally, the review highlights emerging challenges such as recognition in unconstrained environments, post-mortem scenarios, mobile device adaptation, and presentation attack resilience. Results reveal significant progress alongside persistent research gaps, particularly in the areas of cross-domain generalization. This comprehensive analysis provides critical insights and outlines future research directions, guiding scholars and practitioners in advancing iris biometric systems toward greater robustness, scalability, and adaptability across constrained and unconstrained operational environments.