期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2017
卷号：8
期号：3
页码：39-43
语种：English
出版社：Ayushmaan Technologies
摘要：Iris recognition security system is a most versatile research field of recent years. The Iris recognition system comprises of a segmentation system that is based on the Hough Transform Gamma Correction and Histogram Threshold method. The combination of these systems is able to localize the circular Iris outer circle and inner pupil region, reflection, rejecting eyelid and eyelashes. In this paper, we will discuss the different steps to design a system that process an Iris image through image processing steps such as acquisition, segmentation, normalization, feature extraction and matching. However, it is highly probable that images captured at a distance, without user’s cooperation and within highly dynamic capturing environments lead to the appearance of extremely heterogeneous images, with several other types of information in the captured Iris regions (e.g. iris obstructions by eyelids or eyelashes and reflections). The algorithm is implemented over CASIA v4.0 database, IIT Delhi Database and ICE 2005 database and the accuracy of 99.20%, 97.71% and 98.16% is achieved respectively which is seen better from other literature studied and cited in the work..For the implementation of proposed design, the Image Processing Toolbox under MATLAB software is used.
关键词：Iris Recognition;Biometrics;Iris Segmentation;Gamma correction;Histogram Threshold;Growing based segmentation; PCA;Gabor filter;Matching;Codification;Normalization and Image Processing.
