If there is no rest for the wicked, then I have been very evil these past two weeks indeed. But, I hope some of the findings are wickedly cool. This post is for the paper on new findings in security printing and imaging, and my co-authors are that talented trio of Guy Adams, Jason Aronoff and Margaret Sturgill. Though a bit of a grab bag, this paper focuses on three primary tracks of research:

(1) Does error correcting code always make sense for barcodes? We found out that for 2D and 3D barcodes, the answer is, not usually.

(2) Can classification of authentic and counterfeit samples actually improve with increased compression? Interestingly, the answer appears to be yes in many cases, especially when color is involved (I've reviewed a paper recently that showed the same controversial results--increased accuracy with compression--for forest classification).

(3) Does color help forensic image analysis? Again, the answer is yes.

Give it a read, it's short!

http://www.hpl.hp.com/techreports/2009/HPL-2009-328.html

Cheers,

Steve