期刊名称：International Journal of Computer Technology and Applications
电子版ISSN：2229-6093
出版年度：2011
卷号：2
期号：4
页码：884-893
出版社：Technopark Publications
摘要：The Iris pattern is an important biological feature of human body. The recognition of an individual based on irispattern is gaining more popularity due to the uniqueness of the pattern among the people. In this paper PCA based iris recognition using DWT (PIRDWT) is proposed. The upper and lower portion of the iris which is occluded by the eyelids andeyelashes is removed using morphological process. According to Springer Analysis of CASIA data base, to get better recognition forty five pixels to left and right of the pupil boundary is considered as iris template for the proposed algorithm analysis. The image is enhanced using Histogram Equalization to get high contrast. DWT is applied on histograme qualised iris template to get DWT coefficients. The features are extracted from the approximation band of the DWTcoefficients using PCA. Multiple classifiers such as KNN, RFand SVM are used for matching. The proposed algorithm hasbetter performance parameters compared to existing algorithm.
