Cryptography Reference
In-Depth Information
Figure 17.1: The top line shows a picture and its least significant
bits before approximately 6000 bytes were hidden in the image with
EzStego. The image had a capacity of 15000 bytes. The bottom line
shows the image and the least significant bits
afterwards
.
can still be seen in the least significant bits. When the image is either
all white or all black, the least significant bit is far from random.
It's usually pegged at either a zero or a one. Even the nonsaturated
sections are far from random. The positions of the objects in an
image and the lights that illuminate them guarantees that there will
often be gradual changes in colors. These gradients are also far from
random. There's also just pure imperfection. Digital cameras do not
always have a full 24 bits of sensitivity. The software may pump out
24-bit images, but only after padding the results and adding extra
detail. The least significant bits are not always assigned randomly
when the sensors do not have sufficient resolution.
“In the long run, there
are no secrets. in
science. The universe
will not cooperate in a
cover-up.” - Arthur C.
Clarke and Michael P.
Kube-Mcdowell in
The
Trigger
.
After information is hidden in the least significant bits, though,
all of these regions become much more random. The eye is often
the fastest tool for identifying these changes. It's very easy to see the
effects in Figures 17.1 and 17.2.
There is no reason why more complicated visual presentations
can't be created. Information does not need to be hidden in the least
significant bits, in part because it is often very fragile there. [SY98]
More complicated presentations might combine several bit planes
and allow the attacker to try to identify where the extra information
may be hidden.
