As computers are more and more integrated via
networks, the distribution of digital media is
becoming faster, easier, and requiring less effort to
make exact copies. Especially the enormous popularity
of the World Wide Web, starting in the early 1990's,
demonstrated the commercial potential of offering
multimedia resources, such as digital audio, video,
and images, via digital networks. Since commercial
interests seek to use digital networks to offer
digital media for profit, they have a strong interest
in protecting their ownership rights. Digital
watermarking is one of the technologies of copyright
protection for digital video products. In previous
work, all the watermarks were time-invariant;
watermarks are embedded into I-Pictures or P-Picture
blocks with zero motion vectors. Time-variant
watermarks are first proposed in this research. We
first investigate the DCT based time-invariant
invisible watermarking algorithm of I-Pictures and
P-Pictures with non-zero motion vector. Then the
algorithm is extended to yield time-variant
watermarks. With time-variant watermarks, different
video pictures have different watermarks. Different
watermark pictures have different insertion points and
different block mapping methods. This makes our
time-variant watermark more robust than all previous
ones. To resolve rightful ownership, the watermark
extraction requires the original unwatermarked video
data.