Robust video watermarking using significant frame selection based on coefficient difference of lifting wavelet transform

This manuscript presents a significant frame selection (SFS) and quantization of coefficient difference based robust video watermarking scheme in the lifting wavelet transform domain (LWT). In this scheme, a new procedure based on the mathematical relationship between a number of original video frames, coefficient block size and embedding capacity for significant frame selection is proposed. Third level frequency sub-bands of selected frames are obtained using LWT and lower frequency sub-band (i.e. LH3 sub-band) is considered for watermark embedding. Watermark bits are embedded using the quantization of coefficient difference of two maximum frequency components of significant frame blocks. To improve the security against an external intruder, secret key based randomization of video frames, blocks of coefficients and watermark bits is incorporated. Various geometric and image processing operations are tested and the proposed scheme successfully proves its robustness without compromising the quality of watermarked image. Comparing with other existing schemes, a remarkable improvement in terms of robustness is observed however in some cases the imperceptibility is compromised.