Iris biometric security using watermarking and visual cryptography

Abstract

This paper presents a novel security architecture which protects the iris using watermarking and visual cryptography techniques. The iris region is protected using watermarking technique where we are using DCT to embedd the watermarked text image into the iris image in the low frequency band. After performing watermarking, the watermarked image is furthur processed to generate a template using Daugmans method and Gabor filter. The generated template is then gone through another security protection scheme using visual cryptography. Here the template is divided into two shares where one is present with the user and other in the data base. This is the enrollment module. Now for the Authentication module the user share and the data base share are overlapped and then the template is generated. Now the generated template is matched with the template which was already generated using Daugmans method. If both the templates are matched authentication is provided otherwise the person is unauthenticated. This paper is implemented using Matlab.