An Improved Watermarking Algorithm Robust to Camcorder Recording Based on DT-CWT

摘要

To reduce blocking effects in dark and smooth regions, to improve the correct rate of the extracted watermark sequence when frame dropping happens and the video has lots of short shots for existing single-shot watermarking algorithm, and to block against camcorder recording. The watermark sequence is first divided into different subsequences in this paper. By combining the Watson vision model based on Antonini9-7 Wavelet transform and Adaptive embedding intensity coefficient, the watermark subsequence is embedded into the 4th level DT-CWT (Dual Tree Complex Wavelet Transform) low frequency component of the whole frame. To improve the correct rate of the extracted watermark sequence, we use the luminance trend of Y component to extract the time synchronization subsequence and watermark subsequence, and use the maximum election statistics to restore the watermark. Experimental results show that the proposed algorithm can reduce blocking effects, improve the correct rate of the extracted watermark sequence, and is also robust to camcorder recording.