Iris Localization And Recognition Algorithm

With the development of information technology in contemporary society, the demands for security are becoming more and more important. So the technology of biometrics identification has been improved and applied in many fields. Iris Identification is one of potential biometrics technology, it has been gradually applied to safety fields. Iris Identification is based on iris texture to identify persons, and it is one of most accurate biometrics technologies.A personal identification system based on iris pattern is composed of iris image acquisition, image prpprocessing, feature extraction and matching. Iris images for this paper are provided by Institute of Automation Chinese Academy of Science. This paper presents an iris recognition arithmetic which gets a better result in the experiment. The main work of the dissertation is as follows.(1) The drawbacks of some popular iris location algorithms are analyzed firstly, then an improved coarse-to-fine approach is proposed, which reduces the effects of eyelash occlusion and boundary blurring, two major affected factors. With this approach, the inner boundary is quickly located by searching a coarse center and the outer one by image converting, enhancing, and differentiating. The proposed approach is compared with two commonly used ones by experimental results on the CASIA database and is proved its rapidity and precision.(2) In the image proprocessing, the located iris is transformed from an annual area to a rectangular area, then the image is enhanced and converted to some 1D signals which will facilitate the following work.(3) A previous way of iris feature extraction based on Gaussian-Hermite Moment is studied and applied in this paper, and the imfluences of some important parameters are explained.(4) To classify the feature vevtors extracted by Gaussian-Hermite Moment, the drawbacks of the previous way is analysed first, then two new approaches are introduced. One is based on the way of 0-1 coding and Hanming matching, the other is performed with KNR classifier.Experiment results show the effectiveness of this method on CASIA iris data. It is...