Spread transform video watermarking using 3D complex wavelets

A novel video watermarking system is presented that uses a spread transform technique to securely embed data into a video sequence, resistant to both spatial and temporal attacks. The system makes use of the 3-D wavelet domain, improving embedding by extending this to include complex as well as discretely sampled wavelet transforms. To ensure imperceptibility an adaptation of a proven video visual model is used. Results are given for both compression and frame dropping attacks.