SVD-EBP Algorithm for Iris Pattern Recognition

TL;DR

This paper introduces the SVD-EBP system, combining SVD for feature extraction with EBP neural networks for iris image classification, aiming to improve accuracy and efficiency.

Contribution

It presents a novel integrated approach using SVD and EBP for iris recognition, optimizing feature extraction and classification.

Findings

Effective reduction of input vector dimensions using SVD.

Improved classification accuracy with the combined SVD-EBP system.

Demonstrated potential for enhanced iris recognition performance.

Abstract

This paper proposes a neural network approach based on Error Back Propagation (EBP) for classification of different eye images. To reduce the complexity of layered neural network the dimensions of input vectors are optimized using Singular Value Decomposition (SVD). The main of this work is to provide for best method for feature extraction and classification. The details of this combined system named as SVD-EBP system, and results thereof are presented in this paper. Keywords- Singular value decomposition(SVD), Error back Propagation(EBP).