The Design of Software (CLOSED)

A public forum for discussing the design of software, from the user interface to the code architecture. Now closed.

The "Design of Software" discussion group has been merged with the main Joel on Software discussion group.

The archives will remain online indefinitely.

Censored text depixelating algorithm?


I came across an article today which made a passing note that pixelated text can be easily integrated back to its original form. I come across a comment about this every few years and always go on a hunt for a scientific paper or constructive forum discussion about it, but never succeed!

Do you have any suggestions as to where I should be looking for information (algorithms etc) on this? Perhaps I am not phrasing myself correctly - is there a name for this act of depixelating?

Thank you.
Barnie Grumble Send private email
Friday, September 12, 2008
If you are new to topics like adaptive deconvolution you might try starting with a search on Kalman filters.  I'm not sure if Kalman filtering is the best way to go about solving that particular problem but even if it isn't it may still be a more productive search term than depixelization.
system of a Don Send private email
Friday, September 12, 2008
Patrick McKenzie (Bingo Card Creator) Send private email
Sunday, September 14, 2008
Interesting.  He makes an excellent argument for blacking out "sensitive info" instead of blurring it.

Though, having "blurred" info, it's quite possible there's SEVERAL valid values that will result in the same blur, and he doesn't cover that argument very well.

Still, leaving NO information is a superior strategy.

The bottom line for me is that blurring an image should remove sufficient detail (meaning you could de-blur to many possible images) that the original image is effectively not recoverable.
Monday, September 15, 2008
Some image types, such as PNG files, will have several layers of image data.  Sometimes you can create a blacked out area, with the clear image on a lower level of image data.  So it's best to use a format that has only a single level, or ensure that multiple levels are flattened out.
What is this Web stuff, anyway
Monday, September 15, 2008
Or, just for a laugh, replace the real information with fake information - "flattening" it and replacing all the pixels completely - and then blur it. So when they extract the unblurred text, it won't help as much as they hoped.  ;)

And even if it's imperfect recovery, knowing that each digit is most likely one value and could plausible be another is a lot of info. So the easy option is just to overwrite all the pixels with a big black block that doesn't preserve anything.

Completely removing the information is also a good idea when you make a face shot out of a photo of yourself naked while using a format that preserves the full photo's data and just hides the bits you wanted cropped out.  ;)

Monday, September 15, 2008

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics
Powered by FogBugz