A hybrid transforms-based robust video zero-watermarking algorithm for resisting high efficiency video coding compression

Abstract

With the rampancy of pirated videos, video watermarking for copyright protection has become a widely researched topic. In this paper, zero-watermarking is applied to videos for the first time to resist high efficiency video coding compression, which can improve the robustness of the watermarking algorithm and ensure the videos' quality. A robust video zero-watermarking algorithm based on the discrete wavelet transform, the all phase biorthogonal transform, and singular value decomposition is proposed. Utilizing the properties of hybrid transforms, robust features can be extracted from videos, and robust zero-watermarks can be constructed. Experimental results demonstrate that the proposed algorithm has strong robustness to high efficiency video coding compression attacks with different quantization parameters. In addition, the algorithm can also resist common image processing attacks, geometric attacks, frame-based attacks, and hybrid attacks. Compared with existing video watermarking algorithms, the proposed algorithm can more accurately and completely reconstruct watermark images.