USO DE LA RED VASCULAR DE LA RETINA COMO MEDIO BIOMÉTRICO DE IDENTIFICACIÓN

Abstract

The retina has many desirable characteristics as a basis for authentication, including uniqueness, stability, and permanence. This article presents a fast, efficient and automatic retinal-based recognition system. The system is composed of 3 main modules including blood vessel segmentation, feature extraction and feature matching. Sensitivity, specificity and accuracy indexes were computed to establish the performance characteristics of the proposed vessel segmentation method. The obtained indexes are: 0.964855, 0.948025 and 0.975905 respectively. The experimental results show that the proposed segmentation method performs well in extracting vessels. Experimental results on a database, including 200 retinal images obtained from 40 subjects, demonstrated an average equal error rate close to zero and an average accuracy equal to 99.0 percent for the proposed retinal-based recognition system.