Using the stability of the direct current (DC) coefficient values in blocked discrete cosine transformation (DCT), a blind video watermarking method is proposed in the paper. The frame to embed watermark is randomly selected, and its luminance is transformed by 16-by-16 DCT. The watermark is embedded in the low frequency coefficients of the DCT blocks, which are chosen by its DC coefficient values. The watermark is detected by computing the correlation between the watermark and the absolute values of the low frequency coefficients of the blocks selected by DC values. The experimental results demonstrate, whether I, B or P frames are watermarked, the correlation curve emerges peaks in the watermarked frames which suffered compression and decompression with different bit rate. The watermark scheme has strong robustness against unintentional or malicious attacks such as frames deleting, frames inserting, frame statistical average, and collusion attack.
[1]
Ling-Hwei Chen,et al.
A New Digital Watermarking Technique for Video
,
2002,
VISUAL.
[2]
Michael G. Strintzis,et al.
Compressed-domain video watermarking of MPEG streams
,
2002,
Proceedings. IEEE International Conference on Multimedia and Expo.
[3]
Du Shi-pei.
Video Watermarking Technique
,
2004
.
[4]
Ye Lu.
Research of a video digital watermarking algorithm based on MPEG domain
,
2004
.
[5]
Gabriela Csurka,et al.
Robust 3D DFT video watermarking
,
1999,
Electronic Imaging.
[6]
Ahmed H. Tewfik,et al.
Multiresolution scene-based video watermarking using perceptual models
,
1998,
IEEE J. Sel. Areas Commun..