Iris Recognition Based On PKI

Iris recognition is an effective biometrical recognition technology, which will be used widely in information security area like personal identification. In this paper, after extracting the rough edge of the whole image with sobel operator, we locked distinctive continuous areas, considering 9*9 pixies as continuous rule. Then according to the geometrical properties of circle, it will be success to locate inner boundary of iris. Basing on the inner boundary parameters, we use area integral method other than line integral method to locate the outer boundary of iris, which will overcome noise impact in some degree.Gray stretch, gray balance, as well as local statistically methods run to enhance iris image quality during normalization, which adopt bilinear interpolation assist to convert the ring shape of iris to rectangle. We use gabor wavelet transformation to encode texture characteristic of iris, among various other coding strategies, and employ Hamming distance as match criteria.Photographs of iris and those iris images printed on paper, once adopted by the iris recognition system, will cause a lot of trouble. This paper, at first analysis the unique characteristics of live iris and fake iris, then propose an effective algorithm based on FFT, combing image quality assessment method to distinguish them. The experiment result turns out that this feasible algorithm is sufficient to reject more than 95% fake iris images.In practice, we find out that iris recognition system doesn't have ideal credibility in identification or verification mainly because of complicate environment, meanwhile, communication based on PKI(Public Key Infrastructure), doesn't ensure the security of private key storage. To solve this issue, we propose a totally new application based on iris recognition and encryption algorithms, using two hugely prime number generated from the iris code.