A Repetitive Watermarking Scheme for Digital Images based on Self-Inverting Permutations

In this work we present a novel watermarking scheme based on the repetitive application of watermarks inside a digital image. Our approach extends the technique proposed initially in [1] and [2] by implementing the corresponding approach repetitively over the wider area of the digital image and embedding the whole watermark in the corresponding cells instead of parts of it. The approach proposed in this work focuses mainly on securing the watermarked image against crop attacks and compression attacks that constitute the most important attacks through the attack vectors. Our approach provides a significant improvement over the computational cost required for the embed procedure of the watermark inside a digital image, achieving an adequately imperceptible and robust watermarking technique. The experimental results exhibited through the evaluation of our proposed model prove its potentials against these types of attacks and ensure that the embedded information will be successful extracted from the watermarked image after the deployment of such attacks.