Efficient Iris Recognition Using Hybrid Preprocessing and Sheaf Attention Network-Based Segmentation and Classification

Abstract

Iris recognition is a highly accurate biometric identification method that exploits the unique epigenetic patterns present in the human iris. However, many existing approaches encounter limitations related to segmentation precision and classification effectiveness. To address these challenges, this research introduces a novel iris recognition methodology that emphasizes efficient segmentation and classification by integrating convolutional neural networks with Sheaf Attention Networks (CSAN). The primary objective is to design a unified framework that jointly optimizes iris segmentation and classification performance. In this framework, a dense extreme inception multipath guided upsampling network is utilized to achieve precise iris segmentation. Subsequently, various classifiers, including convolutional neural networks enhanced with sheaf attention mechanisms, are evaluated for recognition performance. Experimental results demonstrate that the proposed approach delivers superior accuracy and robustness in iris recognition, making it well-suited for secure authentication and access control applications. Comparative analysis with existing methods shows that CSAN achieves accuracy rates of 99.98%, 99.35%, 99.45%, and 99.65% across four different proposed datasets, respectively..