ABSTRACT
Computerized media content security has progressively turned into an imperative issue. Since video watermarking is a major technology used in copyright protection, a key prerequisite for it is the advancement of the imperceptibility and robustness. We propose a hybrid watermarking scheme for digital videos based on singular value decomposition (SVD) and multilevel Discrete Wavelet Transform (DWT). The two key aspects of watermarking schemes are copyright protection and robustness. In this, we are embedded the watermark in the video frames in two forms using SVD portioning of the watermark, the video frames are portioned into layers (RGB) and DWT sub-band decomposition of host video, for providing copyright protection as well as reliability. Since Scaling factor is an image dependent so we are providing it in the form of a matrix. First level embedding of the singular values is used for substitution of the singular values of the selected video frame layer providing efficient robustness and fidelity of the scheme. The second embedding is used hiding the U and V matrix from SVD of the watermark are converted to binary form and then embedded into the lowest energy band of the DWT transform of the selected layer. Experimental results are provided to illustrate that the proposed scheme is able to withstand a variety of video processing attacks as well as imperceptibility, especially in geometrical attacks.
