A Comparative Study of Reversible Video Watermarking Using Automatic Threshold Adjuster and Non-feedback-Based DE Method

摘要

For authentication and content protection, the act of hiding information in frames is done by reversible video watermarking. An automatic threshold adjuster-based reversible video watermarking (ATAVW) method using difference expansion (DE) for gray-scale video processing system is provided in this paper. To present the best watermarked video frames of an input video stream in terms of peak-signal-to-noise ratio (PSNR), the maximum payload size and the embedded threshold with respect to the embedding capacity is automatically calculated by the proposed algorithm. The main feature of ATAVW method is that the payload size and the embedding threshold are not needed to be specified by the user, like other feedback-based reversible video watermarking algorithms to restore the accurate original video. Software implementation results for an input video having 30 frames are explained based on their quality of services parameters. The ATAVW method is then compared with the non-feedback-based reversible watermarking algorithms. These results clearly demonstrated that the ATAVW method provides higher PSNR and embedding capability with respect to the DE method for reversible watermarking.