A personal identification system based on iris recognition

Abstract

In this paper, a biometric personal identification system based on iris recognition is proposed and presented. The developed system first isolates the iris from the rest of the input image. The separated iris is then normalized and transformed using Wavelets. Classification features are extracted using a novel thresholding technique that keeps track of the relative amplitudes and locations of the extracted high-energy coefficients. An Artificial Neural Network (ANN) is then employed on the extracted features to classify the input image. To show the validity and merits of the proposed system, its performance is compared to that of vector quantization (VQ), a minimum-distance classifier that uses the Euclidean distance. Simulation results show that the proposed ANN system produces a low recognition error of less than 5% and always outperforms the VQ system. The iris images used in this study are obtained from the CASIA database,. © ISCA 2015.