The idea behind the toolkit is that we take a picture, below we take
cove.png, and insert into it a message by changing the colours of the
pixels slightly. Because the changes made are so small, it's hard for
the human eye to see the difference. Because of the way the algorithms
hide it, it should be hard for a computer to tell the difference too.
Benchmarking is provided in the tool to prove this is the case.
are examples of each algorithm run using the 4 least significant bits
of the colour information, and embedding exactly the same file. If you
save the pictures to your hard drive they will be a little bit bigger
than what you see below. Please be patient as this page may take a while
Note: Each of the stego-images below and hiding maps are produced using the "encode" and "simulate" portions of DIIT, respectively.