Scene-based video watermarking for broadcasting systems

In this paper, we propose a new scene based video watermarking method by using the DCT. Scene based and DCT based watermarking give many merits in the view of the processing time and quality of the video. Watermarking embed processing is done by following the variance of each DCT components in the scene-changed frame. Low variance components have a stable status. The comparison of video quality following the embedding parts is done and the results show the improvements. MPEG encoder and decoder are simulated in the desktop PC. Results show the proposed algorithm's effectiveness in real time and usage of a buffer and robustness in the watermarking detection.