Person authentication using multimodal biometrics with watermarking

Abstract

Aim of this project is to implement a novel authentication scheme to establish Digital Rights Management (DRM) based on multimodal biometric verification and watermarking technique. The biometric features that will be used here as iris and face. In this watermark face image is taken to be the host image and the iris feature is selected as the watermark hidden in the host image. Such that iris feature watermark not only protects face biometric data but also can be used as covert recognition. The transformation used for watermark is ridgelet transform. The embedding is performed according to the HVS characteristics. In this regard, we embed data according to two general perceptual rules disturbances are less visible in the highly textured regions, and they are more easily perceived around edges than in textured areas, but less easily than in uniform regions. Having these rules in mind we embed data. In this project first process is the enrollment of the victim's face and iris features in the available database. Second authentication process is done by comparing the features of face image with the features of the face image in the data base when it matches the iris feature is compared with data base if this is also matched then the person is authenticated. This type of biometric provides better authentication and security.