Frequency and Texture Features for Iris Recognition

Digital images and digital image processing have become a vital part of numerous applications, in every day life, science, security, health, etc. The iris of the human eye is a great biometric parameter that can be used for a person’s identification due to its richness and uniqueness in texture and other features. In this paper, a simple method based on the local binary pattern as a texture descriptor and frequency coefficients is proposed. After extracting the eye region, the iris region is found and features are calculated for that region of interest. A support vector machine is used for classification. The proposed method is tested on a well-known CASIA Interval-v4 dataset and the results are improved compared to methods that only use one of these features or a different set of features.