Fast Subsequent Color Iris Matching in large Database

TL;DR

This paper introduces a fast, efficient algorithm for color iris recognition in large databases, utilizing histogram-based primary keys and segmentation techniques to achieve real-time identification.

Contribution

The paper presents a novel algorithm that converts iris image histograms into byte codes as primary keys and segments color iris images for rapid recognition in large datasets.

Findings

Achieves real-time iris recognition in large databases

Uses histogram ratio as primary key for database indexing

Developed and tested in Matlab with high confidence results

Abstract

Databases play an important role in cyber world. It provides authenticity across the globe to the legitimate user. Biometrics is another important tool which recognizes humans using their physical statistics. Biometrics system requires speedy recognition that provides instant and accurate results. Biometric industry is looking for a new algorithm that interacts with biometric system reduces its recognition time while searching its record in large database. We propose a method which provides an appropriate solution for the aforementioned problem. Iris images database could be smart if iris image histogram ratio is used as its primary key. So, we have developed an algorithm that converts image histogram into eight byte code which will be used as primary key of a large database. Second part of this study explains how color iris image recognition can take place. For this a new and efficient algorithm is developed that segments the iris image and performs recognition in much less time. Our research proposes a fast and efficient algorithm that recognizes color irises from large database. We have already implemented this algorithm in Matlab. It provides real-time, high confidence recognition of a person's identity using mathematical analysis of the random patterns that are visible within the iris of an eye.