A Novel Digital Image Watermarking Using Visual Cryptography and Cluster Computing

Digital image watermarking is a technique in which a secret watermark is embedded into the host image to make an assertion about the host image ownership. A practical image watermarking algorithm should be transparent, robust, secure, blind, and fast. Unfortunately, there is hardly any algorithm in the literature with all these characteristics. In this paper we propose a novel algorithm that satisfies all these properties of digital image watermarking. The proposed technique is a grayscale digital image watermarking using novel visual cryptography and cluster computing. In contrast to other algorithms in the literature, the proposed technique constructs two new shares, the size of each share is equal to that of the watermark and their generation is straightforward and fast. The algorithm embeds a given watermark into the Discrete Cosine Transform (DCT) domain. It is robust since it utilizes the relative values of the low frequency components rather than the absolute values. It is transparent since the watermark bits are not physically embedded into the host image. The algorithm is secure due to the use Torus Automorphism (TA) permutation and a pseudo random number generator. It is also blind since few keys and two shares are needed for watermark extraction. Finally, the algorithm relies on cluster computing for significant speedup. The experimental results prove the robustness of the proposed technique against several types of unintentional attacks. Besides, the results prove the practicality of our approach comparing with other techniques in literature.