 Watermarks, inserted in an input video, are
perceptually unnoticable in the output video. VideoMark supports five (5) embedding levels
which are associated to the strength of modifications performed for the watermark
embedding. Generally,
watermarks are detected with high certainty and extremely low false alarm probability.
Independent detection is performed for each frame. Therefore, watermark detection can be
used as a reliable evidence for providing copyright ownership.
Watermarks cannot be extracted or removed without the
knowledge of the corresponding private key. Watermark generation is image (frame)
dependent and cryptographicaly secure.
Watermarks are robust to MPEG2 encoding up to
compression ratios 1:50 (5Mbps compressed video).
Watermarking or detection can be applied to a full
digital video sequence or a section of it.
|
E-mail this page
|
Print this page
Overview