Model-based JPEG Steganography Demo
This demo consists of Matlab routines which implement the JPEG
steganography method described in the paper entitled "Model-Based
Steganography", presented at the
International Workshop on Digital Watermarking in October, 2003.
A preprint of the paper is available
here.
These routines require the Matlab JPEG Toolbox, available
here.
The following Matlab functions are included in this distribution:
arith_decode Non-adaptive arithmetic decoder
arith_encode Non-adaptive arithmetic encoder
blockiness Calculate blockiness statistic for a JPEG image
deblock_delta Compute a delta image that will reduce blockiness
get_sym_frq Obtain symbols and model frequencies for coefficients
get_sym_val Convert symbols to coefficient values
jpeg_steg_capacity Estimate capacity of a JPEG image
jpeg_steg_decode JPEG steganography decoder (used for MB1 and MB2)
jpeg_steg_encode JPEG steganography encoder (MB1)
jpeg_steg_encode_deblock JPEG steganograhpy encoder with deblocking (MB2)
mcdf Model cumulative density function
mcdfinv Model inverse cumulative density function
mfit Maximum likelihood model parameter estimate
mfitc Maximum likelihood model parameter estimate (center bin)
mpdf Model probability density function
mrand Draw samples from model distribution
showimage Display an image at the correct resolution
steg_decode Generic steganography decoder
steg_encode Generic steganography encoder
stegtest Demonstration script for MB1
stegtest_deblock Demonstration script for MB2
This software is free for educational and non-profit use. See the
README
file for licensing information.
Download:
Gzipped tar file (64k, mbsteg_1.2.tar.gz)
PKZip archive (72k, mbsteg_1.2.zip)
README file (6k, README)