Video semantic authentication method based on UCL

In the light of semantic security issue in H.264 video, a strategy of video semantic authentication method based on UCL (Uniform Content Locator) is proposed. In this strategy, UCL information are embedded as semantic watermarks into the medium frequency range of DCT coefficients of I frames and semantic feature information are embedded into motion vectors of P frames. The strategy is implemented by incorporating a UCL-based watermarking method. Experimental results show that this method can effectively embed and protect semantic information under a variety of attacks and it also has little effect on video code rate change and content distortion.