To improve the robustness of the watermarking method remarkably, this paper proposes a video semantics watermarking method, which combines video semantics with video watermarking by using the high-level semantic features of stability and unvulnerability. This method looks upon mining association rules as a constrained optimization problem, uses an improved clonal selection algorithm for the fuzzy association rule mining to extract the semantics of motion and texture, and generates dynamic semantics watermark online. This method determines the shots of interest according to motion semantics adaptively, determines the I frames of interest according to texture semantics adaptively, and selects the intense movement and slow movement of the region as a region of interest according to human visual masking properties, and then embeds watermark in the IF DCT coefficients of the luminance sub-block prediction residual of the I frames of interest; it adaptively controls watermark embedding strength by use of video texture features. Experiments and analysis show that this method is robust not only to various conventional attacks, but also to re-frame, frame cropping, frame deletion and other video-specific attacks.
