|
ABUTALA wrote:
> Implement the spread-spectrum watermarking.
> Your program should embed 16 bits of ‘intelligence information’ in an image of arbitrary size.
> The detector should be able to detect 100% of the embedded bits.
> The report should address the following questions
> a. How does the spreading of message bits occur?
> b. What is the difference between the blind embedding and non-blind embedding?
> c. Why do you need the prefiltering step in blind detection?
> d. Why do you think the whole process work?
> 2. For the above code,
> tweak the gain/strength parameters to come up with a graphical relation
> between perceptibility and detectability of the watermarked image.
> You may use PSNR metric for perceptability and bit error rate for the detectability.
Your posting contains no Matlab questions. This newsgroup is not a
newsgroup in which to discuss image processing techniques in general terms.
It is also not a newsgroup in which to post assignments with the hope that
someone will supply you with the code.
Use your course reference material to locate appropriate algorithms,
and write a trial matlab implementation to implement the algorithms,
and do your best to debug the program. For the first pass, it is best
to worry about getting the program working and getting the right answer:
once you have a working version and understand the techniques involved,
*then* you can worry about making your code more efficient or
"more MATLAB-y".
Get back to us when you have questions that are directly related to Matlab,
such as "How come mod(2^53+1,2) == 0 ?".
If Matlab gives you error messages that you do not understand, then the first
thing that you should do is ask Matlab for more explanation by giving the command
why
at the command prompt.
|