Hybrid quasi-3D DWT/DCT and SVD video watermarking

Abstract

A video watermarking robust algorithm based SVD is proposed. In order to analyze both the spatial and temporal characters of video sequence, each frame of the video is transformed to wavelet domain by 2-D DWT, and then 1-D DCT is used along the temporal axis. Watermarks were embedded into the singular values (SVs) of the coefficient frames and in the extraction procedure watermarks can be detected without the original video or any other information of the original singular values. Experiments show that the watermarked frames are indistinguishable from the original frames subjectively and the algorithm bears desirable robustness on filtering, additive noise, lossy compression, rotation and synchronization attacks etc.