QIM algorithm

[tm022woolfy]

Hi was wondering how much work it would be to implement the Quantization Index Modulation method of steganography in the toolkit ?

I am attempting to do it in a project for uni using Java to insert invisible watermarks into clinical data (only got a month left now I am really stressed out !!!) and it is proving very difficult any help or pointers you could give me would be much appreciated.
Thanks
Tom

[KathrynH]

Hi Tom,

Unfortunately QIM isn't something I've looked at too greatly when I was originally making the toolkit, as far as I know it involves JPEG compression and watermarking, which really puts it outside the scope of where DIIT currently is.  My plan always was to have LSB steganography in the toolkit because the method is simple enough to help people learn about steganography, and JPEG imags can't have that type of steganography (because JPEGs are lossy).  Watermarking are also a problem, since watermarks are generally quite robust and LSB steganography isn't.  You might like to call the steganalytic methods here "weak watermarks" - such as those used to see if an image has been changed.

My suggestion to you if you are happy to use an existing algorithm is to get in touch with the authors of a paper about QIM steganography or steganalysis and see if they are prepared to give you a copy of their algorithm for your project.  Short of re-implementing the algorithm, you probably won't be able to republish their code without their permission, but you could refer to it in an academic report and use it privately.

Good luck.
Kathryn.