How Photoshopped Is That Picture?

Freddybear writes "Digital forensics experts at Dartmouth have developed software that can analyze digital photos to rate how drastically they have been altered by digital editing techniques. 'The Dartmouth research, said Seth Matlins, a former talent agent and marketing executive, could be "hugely important" as a tool for objectively measuring the degree to which photos have been altered.'"

Comments

Re:Too bad this requires a "before" picture

[hedwards]

It's a hard problem, just because a photo looks photoshopped doesn't necessarily mean that it was. These days one can shoot in RAW or TIF which makes the compression artifacts that used to be helpful non-existent. And ultimately somebody that's willing to put the time and effort into the work is probably going to be able to make it so that it fools the software most of the time.

Re:Too bad this requires a "before" picture

[__aavqan3009]

It needs a "before" pic? Really? (could`nt reach the provided link) Could`nt one simply look at the before pic and see if there was retouching? We need software for this?

Re:Too bad this requires a "before" picture

[jandrese]

For some time now there has been resources online to look at the error rates in a JPEG to guess which parts have been photoshopped into a picture. I use it all the time when I suspect something has been photoshopped. It's not a perfect tool,and someone who is expecting it can defeat this particular analysis, but online it has proven to be quite valuable in spotting fakes.

TFA's link appears to be slashdotted, so I can't tell if they're using a similar technique or not.

Re:Too bad this requires a "before" picture

[canajin56]

It's for rating it. Says so in the headline and the summary. So there's an objective number. 1 means "small touch-ups" and 5 means "might as well be a CGI model".

Re:It Can Tell by the Pixels

[kelemvor4]

The whole term is dumb. You don't photoshop something in GIMP, for instance... just like you don't xerox on a Cannon.

http://dictionary.reference.com/browse/xerox

noun 2. ( sometimes lowercase ) a copy made on a xerographic copying machine.
verb (used with object), verb (used without object) 3. ( sometimes lowercase ) to print or reproduce by xerography.

See also: http://dictionary.reference.com/browse/photoshop — vb , -shops , -shopping , -shopped ( tr ) to alter (a digital photograph or other image), using an image editing application, especially Adobe Photoshop
Notice it says especially, not exclusively.


My point is that you most certainly do photoshop something in G.I.M.P..

Re:/. being sued in 3, 2, 1...

[dgatwood]

The best part of that is their suggested replacement for "Photoshopped", which is "enhanced with Adobe® Photoshop® Elements software." This has several problems:

• It contains two registered trademark symbols in the sentence. In addition to making the sentence harder to read, these symbols are only required (legally) if you work for Adobe or are otherwise using the term as part of selling their product or a competing product. It does not dilute Adobe's mark if the term is used in reference to their product, with or without that symbol.
• If you are going to insert those ® marks, then Elements should have a (TM) mark beside it (which curiously, Slashcode does not allow even though it allows ®).
• Such images are rarely actually enhanced by the process. The term "Photoshopped" usually refers to constructing a new image that adds somebody into a picture where they didn't actually appear, makes it look like a bus is falling off a cliff, pastes one person's head on another person's body, etc.
• History has shown that because of the Streisand Effect, guidelines like this are more likely to increase misuse of the mark rather than diminish it.
• Verbing a trademark only significantly dilutes the mark if people start using the word Photoshopped to also mean GIMPed, Pixelmatored, or MS Painted. Since the Streisand Effect applies here anyway, the company would be better off registering a trademark with the verb form, then begin marketing the word Photoshopped as "edited with Photoshop" in the most clear and unambiguous way possible.

As always, caveat emptor. IANALBIPOOSD.

Re:Having a little experience here

[squidflakes]

Most DSLRs have "HDR" capabilities. I know that on Nikon it is called bracketing and you can take three or four images with up to a 2 EV increase or decrease on on either side. That still only gives the image out of your camera 10 - 12 EV of range, which bumps right up against what most monitors are capable of displaying.

Re:Celebrity culture...

[hankwang]

And even before photos are loaded up in Photoshop the celebrity has already been loaded up with a pound of makeup, sat under carefully positioned lights and been photographed by a professional.

Indeed. Dove Evolution video clip

Re:Revert?

[dgatwood]

Creating a Photoshop job that will fool even a sophisticated human eye is easy. Creating one that will fool an algorithm is very, very hard. The modification detection algorithms I remember hearing about start by taking a series of test images with the specific camera that was used to take the picture. In order to beat them, among other things, you'd need to:

• ...mathematically compute the probability of noise for each subpixel and adjust your noise so that the distance of each pixel from the mean of nearby pixels in areas of low contrast is close enough to what would be expected for that particular spot on that particular CCD at a particular physical temperature, and so that the noise level is consistent with the expected noise for a single physical temperature value across the entire image.

  Alternatively, if a particular camera gets hot spots on parts of the chip when shooting lots of pictures in a row, the noise level might need to be a very complex gradient with the hot spots in particular places on the chip.

• ...know where every dead subpixel is on that camera so that you can mathematically compute the correct color channel value for each dead subpixel based on its neighbors in the same way that the camera does.
• ...read the EXIF data to determine which pixels the camera mapped out because of dust, if that particular camera does that, and compute their values programmatically in a similar fashion.
• ...precisely reproduce the chromatic aberrations of the lens at every point in the image.
• ...precisely reproduce the subtle variations in tint at every high contrast edge caused by the relative positions of the subpixels at that particular point on the CCD.

And so on. It's not a case of artistic training. It's a case of spending months modeling a single camera in MatLab. Not a single model. A single camera.

Re:tiff != lossless

[shutdown -p now]

TIFF = Thousands of Incompatible File Formats.