Slashdot Mirror


Interpol Unscrambles Doctored Photo In Manhunt

jackpot777 writes in with an AP story out of Paris reporting that Interpol has distributed photos of a man suspected of sexually exploiting children. The images were recovered from pictures taken off the Internet in which the man's face had been blurred using something like Photoshop's Filter > Distort > Twirl tool. German police were able to recover recognizable images of the man, whose identity and nationality are not known. Interpol would not discuss the techniques used to recover the images. jackpot777 writes: "It does show one interesting facet of internet privacy that has also been noted with topics ranging from reading blurred check numbers in images to Google's plan to blur out license plate and face data for Street View. And that is: blurring is not the same as completely obscuring. As computers become more adept at extrapolating data of different types, your identity isn't safe unless you completely cover all those identifying features."

16 of 370 comments (clear)

  1. Pictures by avij · · Score: 5, Informative

    The pictures can be seen on Interpol's site.

    --

    Follow your Euro bills at EBT
    1. Re:Pictures by petermgreen · · Score: 2, Informative

      almost certainly

      the way I would approach this is to get hold of all major photo editing packages and then do a twirl in a known location on test images. working from those twirled test images it shouldn't be too hard to work out what is being moved to where. If the packages are scriptable this is a lot easier. If not then its a lot of grunt work.

      once you can reverse a twirl in a known location it then just becomes a matter of moving your twirl reversal tool arround until you get a sane looking image.

      --
      note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
    2. Re:Pictures by mikael · · Score: 4, Informative

      A convolution filter would allow you to sharpen or blur the image, or maybe even pick out high frequency detail.

      The spiral distort effect is simply a mathematical function to map one point in a rectangular mesh to another. You basically convert integer pixel coordinates into a floating-point coordinate system with the origin at the centre, apply a rotation based on the distance from the origin, convert back into integer pixel coordinates and transfer the pixel data.

      Consequently, since every pixel is remapped to a new position, the transformation can be reversed.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    3. Re:Pictures by slimak · · Score: 2, Informative

      This is pretty standard stuff in almost any junior/senior level signal or image processing course. Whether the operation can be undone depends heavily on the type of blurring. For example, if the blurring operation is linear then it can be represented by a matrix operation. This means we can write the blurred image (y) in terms of the blurring kernel matrix (H) and the original image (x) as: y=Hx (assume x and y are column vectors of all image samples). This operation can be inverted perfectly as long as H is a non-singular matrix as: x_hat = inv(H)*y = inv(H)*H*x = x.

      When H is singular or non-rectangular the typical approach would be to make x_hat = inv(H^T*H)*H^T*y, which is the least-squares approximation of x from the blurred observation y (^T denotes matrix transpose).

      When the operation is non-linear it generally get MUCH harder. Here, the most common approach would be to use some statistically optimal method or incorporate a priori image information.

  2. Interpol not the ones to descramble by packeteer · · Score: 4, Informative

    From the interpol web page it says:

    These pictures have been produced by specialists from Germany's federal police force, the Bundeskriminalamt, working from originals found on the Internet, which had been digitally altered to disguise the man's face.

    --
    unzip; strip; touch; finger; mount; fsck; more; yes; unmount; sleep
  3. The actual AP article by avij · · Score: 4, Informative

    .. can be read here.

    --

    Follow your Euro bills at EBT
  4. Blurring different from twirling... by Max+Romantschuk · · Score: 5, Informative

    A twirl is essentially shifting pixels around an image, and is designed to keep as much information as possible.

    A blur on the other hand, especially a gaussian blur, will mix pixels together in such a way that any recovered image will be one of many possible outcomes.

    Then again, removing information, by pixellating for example, would be best.

    --
    .: Max Romantschuk :: http://max.romantschuk.fi/
  5. blurring != obscuring; true, but... by Animaether · · Score: 3, Informative

    true, blurring isn't the same as obscuring. That said, a twirl/swirl filter isn't a blur filter either. A twirl/swirl filter relocates pixels from position A to position B. The original pixels are still largely there, you just have to move them back from B to A. That's what Interpol did here - kudos to them for figuring that out. But a blur filter doesn't just relocate pixels - it blends a bunch together. Now don't get me wrong - there's certainly deconvolution methods to reduce blur - especially motion blur - ( one example software: http://www.focusmagic.com/ ) but you're not going to be able to just take any heavily non-motion blurred image and get a supersharp result back. Other techniques, such as pixelization, are even worse to restore - you may as well not try.

    -That- having been said.. yes, obscuring does tend to be better.. as long as it's a proper obscuring and not some half-hearted attempt by a news station where an interviewer / whatever has said to want to be inrecognizable, and then you just get a dark silhouette of the person where you can 1. still make out the silhouette, 2. their voice goes unaltered, 3. bump up the brightness enough and you can even make out a face or, in the case of yea olde license plate, a black bar that is supposed to 'track' the license plate properly, but the person applying the bar is a lazy-ass tracker and it 'swims' over the plate, revealing tiny bits of the bottom/top on certain frames - not too much guesswork involved to figure out the proper license plate, as even with multiple possibilities, only one is likely to match the type/color of the car when looked up on the interwebs.

    Now then... Let the "what if somebody photoshopped somebody else's head on there first, then applied the filter, now some poor innocent sap is framed!" replies begin.

  6. The perfect setup by Anonymous Coward · · Score: 2, Informative

    How soon before someone Photoshops in the face of someone they don't like into KP, obscures it, then releases it knowing the cops will unobscure it and arrest him?

    When he claims "it wasn't me" will he be able to prove it? Sure, the cops are supposed to have the burden of proof but when it comes to bogeyman charges like terrorism and KP juries usually go with the prosecution regardless.

  7. Comment removed by account_deleted · · Score: 4, Informative

    Comment removed based on user account deletion

  8. Re:Super-sekr1t unblurring techniques by Anonymous Coward · · Score: 2, Informative

    In an unbounded picture with infinite intensity resolution, the point spread function of a good old fashioned blur does not destroy information. Then you can recover the image through a process called deconvolution. In the real world, information is lost at the image boundaries and due to requantization, plus you don't know the PSF. If the blur does not cover the whole image and there's only a small number of likely PSFs (as would be the case if the blur was produced in one of the popular image editing programs), two of these deconvolution limitations are already gone and you're only dealing with quantization noise, which isn't so bad if you're only looking for high contrast information like number plate letters. It gets easier if you have multiple pictures or even video.

  9. Re:Once the data's gone, it's gone... by 26199 · · Score: 2, Informative

    It's interesting to note that given enough precision a blur is fully reversable.

    In the frequency domain, a guassian blur reduces the amplitude of high frequencies; it doesn't drop them to zero. Simply multiply up again to get the original image.

    Now, in practice this doesn't work particularly well because the high frequencies end up with such low amplitude that quantization destroys them.

    The reason there is still a potential problem is that the 0-255 precision used on computers holds more information than the human eye can fully perceive. So it may be possible to unblur things when, to a person, the information is gone.

  10. Re:Super-sekr1t unblurring techniques by AmiMoJo · · Score: 3, Informative

    Your method is not entirely foolproof. The problem is that someone can repeat the steps you took and compare the results. They could photograph a person in the same pose, apply the same or very similar filters and if the results match, there is a good chance the source images were also quite similar.

    A similar technique was used to guess blurred out numbers on cheques, passports, car number plates etc. Simply run through all possible combinations of letters and numbers, applying a mosaic each time until the mosaics match.

    --
    const int one = 65536; (Silvermoon, Texture.cs)
    SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
  11. Multiple Images helped Interpol by fast+turtle · · Score: 2, Informative

    Everyone who's talking about blur/twirl/pixelation has forgoten the important thing. Interpol had multiple images. Think of this as having several frames from a film. This gave them enough information to not only un blur/twirl but to compensate for pixelization. So there was a bit more then that used to get the recovered images

    --
    Mod me up/Mod me down: I wont frown as I've no crown
  12. Re:what i've always found crazy about digital imag by Anonymous Coward · · Score: 1, Informative

    While this at first seems like an interesting challenge, a little math will reveal the futility of it:

    Even assuming only a 640x480x256 monitor, you are talking about 307200 pixels -- each of which can have 256 different colors (definately not the photo realism you are looking for).

    307200 ^ 256 = 6.023 x 10 ^ 1404

    This number is by a wide margin larger than the estimated number of atoms in the known universe.
    (4x10^79 according to http://www.madsci.org/posts/archives/oct98/905633072.As.r.html).

  13. Re:In other news... by Punko · · Score: 2, Informative

    If the image is altered prior to the image being declared evidence, then it is not illegal. wiping your fingerprints from a glass is not illegal, unless the authorities have already tagged the glass as evidence.

    --
    If only we could fall into a woman's arms without falling into her hands