An Improved Identification System Using Iris Based on Curvelet Transform and WBCT

Abstract

In recent years, biometric application plays vital role in the security field. Human biometrics such as face, eye, fingerprint, etc is widely used as input for authentication of the person in several authentication related applications. A number of research works are available in the literature on face identification, iris identification, and eye identification. In this paper, a new biometric identification system based on eye especially iris is proposed for authentication. The proposed system consists of three phase’s namely preprocessing, feature extraction, and classification. In first phase, curvelet transform is proposed for image enhancement, after the eye detection and segmentation. In the second phase, wavelet based contourlet transformation is used for feature extraction and best features are selected. Based on the selected features, adaptive neuro-fuzzy inference system is used for classification of the system and it is compared with the existing samples in the database. The authentication is allowed based on the classification results. The simulation result shows that the processing time is reduced and classification accuracy has been improved. The performance of the proposed system has been evaluated through the metrics such as MSE and PSNR values.