Avatar
I'm Drew Breunig and I obsess about technology, media, language, and culture. I live in New York, studied anthropology, and work in advertising technology.

These are reactions to things I feel are important.

Follow me on Twitter.

Fingerprinting Images using Digital Camera Noise 

Every digital picture you snap includes some pixels of noise. Apparently, these elements are unique to your camera and could be used to trace a shot of yours. 33 Bits goes into detail:

First, for each image, a denoising filter is applied, and the denoised image is subtracted from the original to leave only the noise. Next, the noise is averaged across all the reference images — this way the random noise cancels out and leaves the pattern noise.

The takeaway from this seems to be that distinguishing between cameras of different models can be performed with essentially perfect accuracy. Distinguishing between cameras of the same model also seems to have very high accuracy, but it is hard to generalize because of the small sample size.

There’s data everywhere. (Via 33Bits, via PeteSearch)

4 notesShowHide

  1. hopinbitch reblogged this from dbreunig
  2. dbreunig posted this