Steganography: Difference between revisions

From David's Wiki
No edit summary
Line 23: Line 23:
* Postembedding: Embed data into the encoded bitstream
* Postembedding: Embed data into the encoded bitstream


===Methods==
==Methods==
 


==Resources==
==Resources==
* [https://pdfs.semanticscholar.org/e7b1/405d4f344aa0708c76a8f2b1a87c88d21c7c.pdf A Review on Data Hiding Techniques in Compressed Video]
* [https://pdfs.semanticscholar.org/e7b1/405d4f344aa0708c76a8f2b1a87c88d21c7c.pdf A Review on Data Hiding Techniques in Compressed Video]
* [https://www.sciencedirect.com/science/article/pii/S0925231218312608 Video steganography: A review]
* [https://www.sciencedirect.com/science/article/pii/S0925231218312608 Video steganography: A review]

Revision as of 19:59, 10 October 2019

Steganography is hiding data. This can be in audio or videos. The challenge is that the data should be inperceptible to the viewer and recoverable after video compression (e.g. h264).


Requirements

  • Imperceptibility/Impalpability: The video with the embedded data should be visually identical
  • Robustness: The embedded data should not be disturbed by image processing and attacks.
  • Capacity: The amount of embedded data should be high
  • Security: The data should be highly secured. This can be done with encryption beforehand.

The difference between Steganography and Cryptography is the first point. Encryption results in scrambled text which can draw attention.

Attacks include packet loss, video clipping, video scaling.

Classifications

Various ways to classify steganography methods

Embedding Time

  • Preembedding: Embed data into the raw video. Spatial and transformation domains. This will need to survive the encoding process.
  • Intraembedding: Embed data into the video encoding: intra-prediction, motion vectors, pixel interpolation
  • Postembedding: Embed data into the encoded bitstream

Methods

Resources