A Secure Image Watermarking Architecture based on DWT-DCT Domain and Pseudo-Random Number

Abstract

The outburst media contents in today's platforms a need for securing the ownership, authentication and data tampering raised. Protection against tapering the media contents is one of the challenges facing the individuals and industry in a digital age, in another hand rise of artificial neural network models such as deep fake and media contents editors threaten the originality and ownership of media in today digital base life. Demands to secure the ownership and authenticity of data that create by author risen. Hidden the true identity and originality of the media is one way to protect digital creation. Using a watermarked content is one way to protect the ownership of the digital media. It is possible to hide the watermark in the original content to secure it from tapering or illegally usage of the product, by hiding the watermark more securely and acceptable imperceptibility to sensors it is possible to have minimised the effect of watermark process on original contents. In this paper researchers designed an architecture that enhanced and centred on robustness and security of the watermarked image. In this method, we used four random number as seeds generated by a pseudo-random number generator, randomised numbers used as Initial value for encryption of watermark image. Watermark designed in a way that has a lesser footprint on the covered image. Outputs evaluate using PSNR, SSIM and different modification attacks tested against the watermarked image. The result shows that the architecture output is in acceptance range and enhanced the security and robustness of watermarking compere to some existing watermarking architectures and algorithms.