Visual Cryptography with Chaotic Encryption for Biometric Templates

Abstract

Preserving the privacy of digital biometric data (e.g., fingerprint) stored in a central database has become of paramount importance. It demands high speed decryption/encryption process with restricted computational powers. This work explores the possibility of using visual cryptography with chaotic encryption suitable for imparting security to biometric data such as fingerprint images. By using visual cryptography, the original image is decomposed into two images (called shares or sheets) in such a way that the original image can be revealed only when both images are simultaneously available. The security of the stored image can be further enhanced by doing chaotic encryption to the decomposed images. Typically, a private biometric image is dithered into two host images; these images are then independently encrypted using chaotic systems and are then transmitted and stored in two different database servers such that the identity of the private data is not revealed to either server. During the authentication process, Sheets are overlaid (i.e. super imposed) in order to reconstruct the private image. This work proposes a method for ensuring higher level of security for the images using visual cryptography with chaotic encryption. In the first phase of the project, the use of visual cryptography is explored to preserve the privacy of biometric data. This work is being done using Matlab 2007