Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 8 (2022) > Article

DOI: 10.14704/nq.2022.20.8.NQ44724

Machine Learning application for Smartphone Captured Iris Recognition using Image Embedding for Feature Extraction

Vamsidhar.A, Surya Kavitha.T, Supreeth.M, DivyaKrishna.M, VinodKumar.K, PriyaSamhitha.N


This paper exhibits how the machine learning techniques are applied on the iris images which are captured by smartphones. The process begins with the use of Daugman’s approach to pinpoint the location of the iris and the eyelids are subsequently suppressed using the clever edge detection approach. The retrieved iris will then be normalized by choosing an adaptive threshold. Afterward, to procure the feature vectors, an image embedding process named Visual Geometry Group (VGG) is activated to the iris images before processing to the machine learning (ML) algorithm. The VGG architecture is selected in two variants namely: VGG-16 and VGG-19. Both the techniques are repeated for number of iris images and the collected data is given to various machine learning algorithms. The data collected is segregated into two sets, majority for the training of ML algorithm and the rest for the testing. By considering the precision of identification, the efficiency of the algorithm is calculated. Five ML methods are used in this manuscript, and out of all, Logistic Regression has made a significant classification accuracy of 93.33% for VGG-16 and 97.78% for VGG-19 over its counterparts.


Irisrecognition,Daugman’smethod,Visual Geometry Group, MachineLearning, Logistic Regression

Full Text