[WM]: Moment normalization+ robustness to 'photoshop' changes-LOOKING FOR REFERENCES

["Kourosh Fallah" <kourosh@ics.mq.edu.au>]

Hi all,

 

Could anybody please help me about the following questions? References are even more appreciated.

 

1.	Some of the watermarking algorithm use 'moment normalization' to
return the geometrically attacked picture to its original or canonical form.
Assume we crop the picture considerably (say more than 15%), apply affine transforms and then use moment normalization. Will the
final result be the cropped version of the original/canonical picture?
2.	I understand that journalists and digital artist use image
manipulation software, e.g. Adobe Photoshop, to improve the picture appearance or change it to their desired state (e.g. shaping up
celebrities, moving objects, etc). I'm not sure if the various techniques used can be modelled by known random local alterations,
i.e. Stirmark, blind pattern matching, etc. Anybody knows how- if possible- these kind of 'perceptual attacks' can be modelled by
known watermarking attacks or in general how they affect the performance of current watermarking algorithms?



Thanks,

Kourosh

 



______________________________________________________________________________

Watermarking Mailing List - http://www.watermarkingworld.org/ml.html
To unsubscribe send email to "majordomo@watermarkingworld.org" with
"unsubscribe watermarking YOURMAIL" in the body.
______________________________________________________________________________