Video watermarking in the 3D-DWT domain using quantization-based methods
Patrizio Campisi · 2005
Video watermarking methods operating in the three-dimensional discrete wavelet transform (3D-DWT) domain using quantization-based embedding techniques are here proposed. Specifically, the video sequence is partitioned into spatio-temporal units of fixed length. Then the video shots undergo a three level 3D-DWT. Two embedding methods are here considered: quantization index modulation and rational dither modulation. Finally the inverse 3D-DWT is performed thus obtained the marked video shot. The effectiveness of the proposed methods has been verified experimentally. They guarantee a high mark imperceptibility as well as robustness to attacks such as MPEG2 compression, MPEG4 compression, collusion, transcoding, and frame dropping. Gain attack is also considered