Real time implementation of integer DCT based video watermarking architecture

With the recent development in multimedia communication network, data integrity and security of original content is the area of concern. Video is the one of the most popular object which is being shared easily throughout the media. Video watermarking is the current state of research to resolve the video ownership and authenticity related issues. There is a substantial amount of development in software based video watermarking from last few years. The prior works mainly focused on video watermarking that targeted for raw video where the watermark is embedded on the uncompressed video. At the present video capturing devices produce their output in one of the video compression standard. Software watermarking introduces a measurable quantity of delay between video capturing and watermark embedding process. Thus, software watermarking is not one of the ideal choices for real time watermark embedding. In the paper, a novel invisible and robust integer Discrete Cosine Transform (DCT) based video watermarking has been proposed. The proposed video watermarking is developed for real time watermark embedding and can easily be adapted as primary part of H.264 encoder. The proposed algorithm has an essential part in form of integer DCT. Integer DCT is implemented with two different approaches, one is with fully pipeline architecture and the other is recursive architecture, for better speed and area optimization. The robustness of the algorithm has been improved against some video attacks with introducing the concept of scene change detection.