Computationally Efficient and Secure HVS Based Composite Fingerprinting Scheme Using SHS and ECDSA

Abstract

Fingerprinting is an extension to the watermarking principle where each copy of the image is embedded with unique purchaser’s information. Secure fingerprinting aims to prevent efforts to identify the embedded contents. It offers an effective tracing contrivance and protects the legitimacy of digital data. Only an authenticated user can retrieve the implanted contents. In this chapter a secure Human Visual System (HVS) based fingerprinting scheme is proposed for embedding unique composite fingerprints to safeguard buyer’s certification accompanied by image integrity. Combination of authentication and reliability issues is determined using Secure Hash Standard (SHS): Secure Hash Algorithm (SHA-512/256) and Elliptic Curve Digital Signature Algorithm (ECDSA). SHS generates a unique authentication code by integrating image information with buyer’s credentials. Uniqueness property of digest allows for identification of authentication deceit by ensuring maximum collision resistance strength. Composite fingerprint (CF) created using ECDSA is embedded using a maximally separated imperceptible mode, by means of HVS approach. The positions take advantage of color content of digital images to reduce the visible distortion introduced by embedding CF. Signing and verifying digital signatures are two core elements adopted while embedding and retrieving of CF. Test cases are generated for different image dimensions used for embedding varying size fingerprints. Experimental results prove the approach to be beneficial by forming a maximally resilient counter to collusion attacks. An efficient imperceptible semi-blind fingerprinting scheme is achieved without degrading the quality and integrity of the host image.