New algorithm for localization of iris recognition using deep learning neural networks

Ekbal Hussein Ali, Hanadi Abbas Jaber, Nahida Naji Kadhim


Iris recognition is the most reliable and accurate method for eye identification. A novel strategy for localizing iris printing is proposed in this paper. The median filter and histogram were used for this purpose. To extract iris features from iris photographs, an algebraic method known as semi-discrete matrix decomposition (SDD) is used. For classification, neural network (NN) is used to extract the SDD feature. This study also included the setup of convolution neural network (CNN), a convolution neural network that does not require feature extraction, as well as a comparison of the two types of classifiers is made. Iris images are obtained from the Chinese Academy of Sciences Institute of Automation dataset (CASIA Iris-V1), a common database used for the iris recognition system. The proposed algorithm is straightforward, simple, efficient, and fast. The experimental results showed that the proposed algorithm achieved high classification accuracy of approximately 95.5% and 95% for CNN and NN based on SDD features respectively. The proposed algorithms outperformed literature works and required less time for determining the location of iris region.


Biometric identification; CASIA Iris-V1 dataset; Convolutional neural network; Iris image recognition; Median filter; Semi-discrete matrix decomposition features

Full Text:




Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

shopify stats IJEECS visitor statistics