Iris Authentication Using Adaptive Neuro-Fuzzy Inference System

Abstract

As the worldwide need for information security and security laws develops, biometric identification technology is becoming more incorporated into our daily lives. Combining biometrics with encryption has the potential to boost trust in the legitimate data carrier. The generation of keys is a critical topic in cryptographic systems. As a result, the integration of biometrics with cryptographic "keys" is often associated with the application of biometrics in cryptographic systems (or bio-cryptosystems). In this article, we proposed iris authentication using Adaptive Neuro-Fuzzy Inference System (ANFIS) method. The iris recognition is widely regarded as one of the most trustworthy techniques available; we used this metric to determine the authorization of individuals to access server-based data. The ANFIS model for iris scans was developed to identify and classify approved and unauthorized individuals. The ANFIS method was also used to get the key from the characteristics of the iris images. The iris image has enhanced using WFI-CLAHE used. The segmentation has done with Hough transform with Morphological Operation. The feature extraction has done with PCA, LDA, mPCA, SpPCA, LBP models. Finally the IRIS authenticated by ANFIS model and test results reveals that the proposed model is very accurate, far over 96%, and resistant to attacks on server-based data.