Iris Recognition System Algorithm Research

With the rapid development of economy and the advancement of science and technology, people begin to pay more attention to improve identity authentication technology in modern society. In order to improve the low anti-false capability of traditional identity authentication technology, the way of recognizing identity by physical characteristics of human body, such as fingerprint, iris and voice, has been put forward. Thanks to its high reliability and other good features, iris-based identity authentication technology has become the research hotspots of Biometric Identification Technology.Iris-based identity authentication technology is mainly composed of iris image acquiring, image preprocessing, feature extraction as well as matching and recognizing. This article focuses on iris recognition algorithm which is mainly used in preprocessing, feature extraction and mode matching.Because of the slow speed of localization and inaccurate localization of outer-edge, the localization arithmetic is improved. According to the improved arithmetic, the localization of inner circle should include coarse localization and accurate localization. Improvement is made in evaluating the threshold of binaryzation by histogram based on the gray information, also, the center and radius of the circle are localized coarsely by geometric projection in order to reduce the searching scope and calculating amount of arithmetic operator checking of moving circle in accurate localization. Use the improved method of Canny operator combining with Hough transform in localizing out circle.A new approach is adopted for feature extraction in this article. This approach includes the following aspects: the Radon transform of iris image; moment feature matrix obtained from K-order moment based on Radon transform; SVD vector obtained from singular value decomposition of moment feature matrix; coding of the SVD vector as the image feature.At last, it's suggested that the improved Euclidean distance discrimination be used in comparison and identification of iris.The experimental results in CASIA2.0 iris database show that the method of this paper enhanced the speed and accuracy of recognition.