Hide information in video

MSU Graphics & Media Lab (Video Group)

MSU StegoVideo allows to hide any file in a video sequence before compression and extract it after decompression with unknown codec.

When the program was created, different popular codecs were analyzed and an algorithm was chosen which provides the smallest data loss after video compression. Convolutional codes with Viterbi decoding are used to correct arising errors.

The main project goals:

Small video distortions after hiding info
It is possible to extract info after video was compressed
Information is protected with password

Examples

This is an example of extracted info.
# denotes symbol with mistake. In real file you will see any symbol in this place.

Data hiding, a form of
steganography, embeds
data into digital media for
the purpose of identification,
annotation, and copyright.

Source text

Data hiding, a form of
steg#####aphy, e#beds
d#ta into digital #edia for
th# purpose of identificat###,
a####ation, and cop#ri###.

Source text hidden in a film that was compressed by DivX
bitrate 700, framesize 512x384, data redundancy 6

Data##i##ng, a form of
steganography, embeds
data into digital media for
the #urpose of identification,
annotation, and copyright.

Source text hidden in a film that was compressed by DivX
bitrate 700, framesize 512x384, data redudancy 9.
The number of errors decreases when you increase redundancy.

This is an example of distortions in video with hidden info.


source frame	source frame with hidden info

compressed frame with hidden info - information was completely recovered!

Quality of information extracted from compressed video depends on codec, data redundancy and frame size. The greater information redundancy, the greater probability of extracting file info without errors. Generally, the algorithm is applicable at bit rates down to 500 kbps.

Download

MSU StegoVideo as independent .exe(ZIP, 129 KB)
MSU StegoVideo as VirtualDub filter(ZIP, 57 KB)
Unpack this file in the folder "Plugins" in VirtualDub directory to use.

The project team

Project lead:

Dr. Dmitriy Vatolin

Researchers:

Oleg Petrov

Contacts

e-mail:
video@graphics.cs.msu.ru

��

��

��

��

��

��