Syllabus
Digital image formats (raster, palette and JPEG format).
Digital image acquisition (Bayer filter, image processing, noise).
LSB embedding and ±1 embedding.
Histogram attack on LSB embedding.
Quantitative attack on Jsteg.
Sample pairs analysis.
Steganography in palette images, optimal-parity embedding.
Attack on embedding with optimal parity assignment.
Model-preserving steganography (OutGuess embedding algorithm).
F5 embedding algorithm.
Embedding while reducing color depth and its efficiency.
Matrix embedding and its efficiency.
Matrix embedding algorithm using a minimum-distance decoder.
Lower bound on the covering radius depending on relative payload.
Perfect codes and the upper bound on embedding efficiency.
Sum and difference covering sets.
Wet paper theorem, perturbed quantization, embedding with double JPEG compression, double-layered ±1 embedding.
Embedding using the Viterbi algorithm.
Annotation
The course is concerned with the basic notions of steganography in the context of standard image file formats.
The course may not be taught every academic year.