News for hackers. Stuff that matters.
Re:Visual Cryptography by Groxx on Monday June 28, 2010 @06:42PM (#1469694)
[side topic]

The last-linked "additional information" mentions rotated 1/2-circles with black & white halves to define greyscale images. There's an easier way, and it'll scale to as many layers and colors as you want:

pixel(r,g,b) = ((r1^r2), (g1^g2), (b1^b2)). (bit-level xoring) More layers just add in their values. Until you add all of them, there's no way to know what the final value for a pixel will be.

As for the "general access structures"... maybe this can be tweaked, haven't looked closely enough and my brain is tired.

---

edit: fixed algorithm. wrote quick ruby script. first attempt at something like this, glad it works.

edit again: and for those interested, the last linked "additional info" article describes how to make multi-layer images like the one in the main article. It's kinda interesting.

Re:Visual Cryptography by kragen on Monday June 28, 2010 @07:01PM (#1469747)
You can avoid the waste of space implied by the circles by using ½×1 rectangles instead: http://lists.canonical.org/pipermail/kragen-hacks/2007-March...

The interesting thing about the original technique (to me) is that you can do the decryption without a computer. If you want to do this in color, I think you need to use the LCD approach of having red, green, and blue subpixels next to each other.