A Robust Video Watermarking Scheme to Resist Screen Recording Via Codec Simulation

Digital watermarking is crucial for safeguarding video copyrights. However, existing techniques struggle to counter the challenges posed by screen recording attacks, which hinders robust performance. These challenges stem from two main aspects: the non-differentiable hybrid distortion in screen recording and distortion from multiple video codecs during recording and streaming. In this study, we propose a robust video watermarking scheme to tackle these challenges via codec and distortion simulation. To address the first challenge, we model the most influential distortion during screen recording and construct a screen recording distortion layer to combat noise effectively. For the second challenge, we use a video codec simulator to manage distortion from various codecs. Our experiments demonstrate over 92% watermark extraction accuracy for recorded videos under different attacks and achieves more than 99 % accuracy in frame average, frame swap, Gaussian noise and Gaussian blur.