Video watermarking technique using motion vectors to video authentication

With the increase in the number of communication networks and digital multimedia technologies, ensuring the confidentiality of information has become increasingly important. Digital watermarking and scrambling are two methods that can be used to provide security to digital data such as video and to limit the chances of piracy. By implementing these techniques, it becomes more difficult for unauthorized users to access or distribute the video content, thereby protecting the integrity of the video data. This paper presents a novel approach to video watermarking, where a watermark bit is inserted into the motion vectors obtained through motion estimation. This is done to ensure copyright protection for the video content. In the proposed architecture, the negative horizontal components of the motion vectors generated by the three step search (TSS) algorithm are used, followed by a perturbation of the motion vectors (MVs). The perturbation helps to increase the security of the watermark and prevent it from being easily detected or removed. The experimental results of the proposed architecture have shown that it uses few resources, thereby making it a practical and efficient solution for video watermarking. Overall, the proposed architecture provides a better quality watermarked image sequence, operating with low power consumption that does not exceed 1.39 W and a maximum frequency of approximately 75.7 MHz.