Picture of Avoiding Camera Noise Signatures
If you take enough images with your digital camera, they can all be compared together and a unique signature can be determined. This means that even when you think that you are posting a photo anonymously to the internet, you are actually providing clues for the government to better tell who you are. The larger the sample size of images they have, the easier it is them to track down images coming from the same camera. Once they know all the images are coming from the same camera, all they then have to do is find that camera and take a picture to confirm it beyond a reasonable doubt.

It is important to remove this noise signature so that you cannot be tracked down. I cannot guarantee any of these methods will work beyond the shadow of a doubt because the woman doing research for the government on how to find the signature is very good. Check out her papers if you're really good at dense math, and pass along what you learn. I can only promise that this will make their work more difficult.

Step 1: Remove the camera info

Picture of Remove the camera info
When you take an image with a digital camera, the camera itself leaves a data file which automatically identifies itself. This file is called an EXIF and usually has information like the make of the camera, the ISO setting, the date and time, the pixel setting, etc. This needs to be removed. The easiest way to remove this is to open photoshop and save it for web.

When you save, I recommend using a JPG compression with a quality setting of no better than 60%. This will add a lot of ugly noise make it slightly harder to create a noise signature.

The beautiful swiss landscape was found here. It is licensed with a creative commons attribution-share alike 2.5 generic license.
From reading the .PDF file from your source there, she mentioned that image analysis has been performed on analog (film) cameras long before digital image analysis. Therefore, step 7 really won't give you a lot of options.
u_eco4 years ago
This is going to be a bit technical, though can be written in a short python script using Python Imagimg Library:

You can add random noise using some good pseudo-random generator, like Mersenne Twister (seed with something random) - choose random number of pixels at random positions and brighten/darken them a bit (or mess with both chroma and luma channels).

Then use either nearest-neighbor or lanczos interpolation to resize image (instead of bicubic), they modify the frequency spectrum and histogram characteristics much more.

The reason jpeg compression does the described degradation of noise is because jpeg quantization tables used for lower qualities used after discrete cosine transform basically remove highest frequencies ("the noise").
u_eco4 years ago
I suggest either nearest neighbour or lanczos interpolation when scaling images. Lanczos is based on "ideal" sinc filter and can be configured as low-pass filter that removes noise (the statistical signature).

For example, using lanczos interpolation when scaling image in GIMP, it screws up higher frequencies (the visible noise) quite considerably (measured with power spectral density and correlated to original image).

See attached example measurement (shows PSD and correlation of frequencies spectrum after Fourier transform).
u_eco u_eco4 years ago
Here is direct link to image showing PSD and correlation, since instructables downsized it.
elderbear4 years ago
If the government is going to go through this much work to prove you took a bunch of pictures, you're either a terrorist, a spy, or a kiddie pornographer.

What this instructable lacks is a demonstration. A fairly simple one would be to take 4n pictures, where n is a big number, say 1,000 or 100,000, with the same camera in randomly different situations. Do not process the pictures. Create four composite images, summing the pixels of n pictures together (stacking them).

Now process them. Subtract the mean value of each stack from that stack. Compute the standard deviation of all the pixels in a stack and divide the pixel values by the standard deviation.

If the four processed stacks look similar, there is a strong signature in the image. If not, look for a more subtle signature. Calculate a spatial power spectrum from each processed stack. Similar power spectra MAY indicate a signature.

Do a spatial correlation between each of the four stacked images. Similar correlations can also indicate a signature.

If you find a "signature," reduce n by a factor of two and compare the 8 stacks. See how many images it takes to create a signature. Figure the government needs fewer pictures than you do, but probably not a lot fewer.

How many of your pictures would they have to suspect to be from the same camera in order to get a signature?

Just for kicks, do the same thing with a friend's camera of a different model and make. If it has a similar signature ...
The "government" does not have to do anything on such a large scale as ElderBear says. (No insult intended to you ElderBear!)

GOOGLE already has the processing power to do this kind of comparison; and they have the photographic database to test it with. Just as a phone number can be found in many cases via THE GOOGLE, photographs can be examined as well. It is all just data of one kind or another. Data is messy and leaves a trail of one kind or another for someone or something with a trained eye for it.

w1n5t0n: Thank you for writing such a clear Instructable. Your style of presentation was very neat and well presented.
cfuse elderbear4 years ago
"If the government is going to go through this much work to prove you took a bunch of pictures, you're either a terrorist, a spy, or a kiddie pornographer."

Not everyone has the same faith in government that you do. Many don't find an appeal to safety argument to be a convincing one. I certainly don't.

Tell me how being able to identify the noise pattern on a camera wouldn't also allow forged images? That is one potential abuse of this technology (which, given how useful it would be for law enforcement, and how trivial it would be to automate, will certainly end up being used in the courts). I don't think it is difficult to think of scenarios where this kind of identification could be problematic.

In terms of the practical demonstration you mention, I would suggest that it would be better to take your sample images with the lens cap on (so only noise, and no real image data). Additionally, it would probably be wise to re-run the tests several times in a range of (operating) temperatures (given the effect on CCDs it has).

As for the number of images required for identification, that number drops drastically with a CCD noise database at point of manufacture, or with deliberately modified CCDs (or camera firmware) - both techniques would be simple to implement and significantly reduce the samples required to identify.
elderbear4 years ago
This isn't actually pseudo random noise. It is random noise that MAY be biased by process variation between elements of the CCD.
Darwinfish4 years ago
There's nothing wrong with knowing how to do this. (As MahavishnuMan pointed out, information has no intent.) There's nothing wrong with the Instructable. I'm probably taking all this way too seriously. But the paranoia that often fuels this sort of thing can be more dangerous than the perceived threat.

Let's say someone gets shot. The cops look at two guns, both of which match the bullet found in the corpse. Both owners claim innocence. The first gun is unaltered, with the owner's fingerprints where you'd expect them. The second is wiped clean of prints, and the serial number is filed off. Guess who's getting arrested.

The same goes for computers. Say you've got a picture on your hard drive of this amazingly huge sparkler bomb your friend made for the fourth of july. Airport security sees fit to search your hard drive, and they find it. They ask you about it. You cooperate, answer their questions, and assure them that it was all on the level. I'm not saying they'll compliment your friend's bomb making skills and send you on your way, but the point is, they won't have sufficient grounds to suspect you of plotting an attack.

If, however, the file is hidden, sanitized, distorted, and downloaded through a proxy server, and when asked about it you lie, or clam up and demand to see a lawyer, you've just earned yourself an FBI file and a body cavity search. That kind of forethought indicates that you were deliberately hiding something, and that's the sort of thing that law enforcement and airport security focus on. The more sophisticated your methods, the more interested they'll be (if, of course, your concealment efforts fail). Sure, they're wasting their time on you, but they wouldn't be doing their jobs if they didn't investigate someone who is clearly hiding something.

I'm not saying that we shouldn't have some degree of privacy. I don't want to take the airport security guys on a guided tour of my hard drive any more than I want to install a breathalyzer ignition lockout in my car, or put a tracking chip in my arm. My point is that in all but the most exceptional circumstances, this kind of tampering is more trouble than it's worth. Unless we have something to hide, being paranoid only makes others paranoid about us. We cannot live in fear of one another.

Personally, I make a point to not put anything online that I wouldn't put my name on. If "they" (the fearful entity I've heard so much about) do in fact come for me, well, I'll deal with it. Better to die on your feet than live on your knees, right?
Buzzinski5 years ago
Bit narrow minded of some members who seem to think that the US is the beginning and end of the entire world when it comes to things like rights and that any system different to yours must be somehow flawed.

It isn't.

I think that the "Right of free speech' needs to be put into context here . You may have a right to free speech in the USA but in Australia and most of the rest of the western world you do not have this 'right'. You instead are not stopped from voicing your opinion as there is no legal means of stopping you from voicing your opnion. There is a difference.

We have compulsory electrol registration and voting for all citizens aged 18 and over who are not in jail for longer than 6 months or classified as mentaly incompetent to vote etc.  It is also legal to drink alcohol, get married, own property, sign contracts, be called for jury duty at 18 etc as well.
I have to disagree with you. All the democratic countries have the right of free speech some way or another included in their constitution, just like the US. In many times it simple has to be included in the constitution or every time a new government would come they could change the law with plain majority and start jailing people for their opinions. Having it declared in the constitution makes it a tad harder....
plexy CoolKoon4 years ago
Unfortunately Australia has no constitution. To the best of my knowledge nobody has been sent to gaol so far for voicing an opinion in Australia, but freedom of speech is not a right held by Australian citizens and there is no actual law about it. Suing for defamation works somewhat better in Australia that it does in the US.

Yes, I agree that this is a problem and things could become a lot worse, but that's the way it is right now.
Buzzinski plexy4 years ago
http://australianpolitics.com/articles/constitution/full-text-of-the-australian-constitution-as-amended A full copy of Australia's non existent constitution can be found by clicking on the link above Plexy.

http://www.aph.gov.au/LIBRARY/pubs/rn/2001-02/02rn42.htm Section of the non existant constitution pertaining to freedom of speach.

Freedom of Speech and the Constitution
The Australian Constitution does not have any express provision relating to freedom of speech. In theory, therefore, the Commonwealth Parliament may restrict or censor speech through censorship legislation or other laws, as long as they are otherwise within constitutional power. The Constitution consists mainly of provisions relating to the structure of the Commonwealth Parliament, executive government and the federal judicial system.(6) There is no list of personal rights or freedoms which may be enforced in the courts. There are however some provisions relating to personal rights such as the right to trial by jury (section 80), and the right to freedom of religion (section 116).

I couldn't help but weigh in after reading the responses to this interesting (and informative) Instructable.

Over the past decade, the attitudes of both government and citizen have changed considerably with regards to how comfortable we are with perceived threats, as well as how comfortable we can become with opening our lives to outside inspection in an attempt to counter these threats.  Of course, we all know what got the ball rolling and when, and I can't think of a single person in the world that wasn't rocked to their core.  When our hearts sank into piles of smoldering debris, we were all forced to reevaluate how safe we thought we were, and what we would be willing to do to recover the tattered shred of naive security we wore every day.

Fear is instinct - it is a reaction that sharpens our wits and keep us alive as it had for millennia.  But, fear can dull our senses with a single hammer-blow, sending us careening into a wild stampede through the clearest path of escape, leading us eventually over a cliff to our demise with the rest of the herd.

We were instructed as children that we were born free, and had inalienable rights with laws and government to protect them.  After this event, we failed to ask ourselves how the changes taking place in our laws and government stand in direct conflict with the core values we had been taught.  We saw these measures as a means to bring swift justice to the guilty, an instrument that would wrest the cloak of liberty from parties we felt deserved no liberty.

What we didn't realize then, and few realize today, is that the same legal vehicle introduced to stamp out terrorist acts also allow the government the reach to monitor everyone indiscriminately; regardless of whether anyone agrees that their government chooses to act on this ability, the potential exists nonetheless.

And, as any history scholar can tell you, governmental power so absolute corrupts absolutely.

For any readers who view this as scaremongering, remember the fevered pitch under which you chanted "I have nothing to hide" just nine years ago.

Technology's reach and influence in our daily lives has rapidly increased since then.  Furthermore, we have embraced the ubiquity of it as a positive change - a way to broaden our horizons and shrink our borders.  We share pictures on Flickr, then Twitter about them, and get on Facebook to tell all our friends about it.  We freely offer this information to anyone with an internet connection, and if a suspicious party with the authority takes interest in us, they can use this information we volunteer for their own interests.  This may seem Orwellian, but it's not - we, in this sense, are our own Big Brother.

It is no secret that this information exists, nor is it a secret that the government, utilities, insurance companies, and even your employer use this information on a daily basis for thousands of reasons.  When we hear someone warning us of these reasons that sound like science fiction and are far removed from the intended use of the information, we scoff and dismiss it as paranoia.

We often forget that information itself has no intent.  It's no secret that, while Social Security numbers were never intended to be used as personally identifiable, you can't live in the US beyond a cardboard box without one.  But now if you post a happy picture of you smiling on vacation, you could lose your insurance.  If you have a state-issued photo ID, your face is currently being matched against known criminals and suspects.  The only question now remains, what constitutes a crime anymore - and could you be a suspect?

After that horrific day, a new act was hurriedly signed; not even those who signed it read it or understood its contents.  If such an act, one that usurps the very rights granted to all citizens in the interest of protecting the innocent, can slip under the radar, then you must ask yourself what else can be signed into law.

As of 2001, the means exist to hold you indefinitely without a lawyer for questioning over charges which do not have to be made clear to you or anyone else.  Under this same act, all criminals (no matter how small the crime) are a potential threat and are within the scope of suspicion.  All anyone needs is a reason to do so, and some sort of evidence to put you in the wrong place at the wrong time.

But don't worry - you have nothing to hide.
twocvbloke5 years ago
I find it funny, tracing people's movements through images posted online would be going a bit too far, unless it was for dealing with criminal acts, cos the average person, such as myself, couldn't care less if I'm being profiled by some secretive government body, some creepy organisation or an online stalker, but that said, I always crop my images, cos my camera isn't that great... :P

Still, it's a very interesting read about what cameras do with the information and how they store it, might explain why my photos are about 2MB for a single 1280x1024 JPEG picture... :S
darus677 years ago
It may not happen where you live, but there are places in the world where the government gets very uptight about images of some of their activities becoming public. Not everyone seeking anonymity is paranoid.
error23 darus677 years ago
darus67 is dead spot on. The argument of "If you aren't doing anything wrong, you'd have nothing to hide" is insidious and shameful. We all, as human beings, have a right to privacy and free speech - but, in a world where one's every move is recorded, tracked, collated, quantified, and reported (and we've seen proof positive that it is, at least in places like London - I' m sure many other parts of the world are the same), then the only way to exercise these rights is through so-called paranoia. As they quote so often on the Internet these days, "If you give me six lines written by the most honest man, I will find something in them to hang him." Thanks, darus67!
gmoon error237 years ago
Regarding the paper-thin 'I've Got Nothing to Hide' argument against privacy, here's an excellent article from the San Diego Law Review.
What's "Shameful and insidious" is that some people insist on ABUSING the right to privacy. Your so called right to privacy is GONE, it was stolen in little pieces by every criminal or terrorist you've ever heard of, and a lot you haven't. It's sad, but it's a fact. Accept it, live with it, or take it up with the people who use YOUR "rights" to conceal their own shameful and insidious agendas. For my two penny worth, anything that identifies an image as MY property is a good thing. Anything that helps identify a criminal or prove the innocence of a law abiding citizen should be encouraged, and anyone who is researching additional ways to fight that particular battle is a hero. I did find this Instructable very interesting however and I'm grateful to the original poster for bringing it to my attention.
and anyone who is researching additional ways to fight that particular battle is a hero.

I really think that's the point of this instructable--that the "battle" has far overshot it's legitimate anti-terrorism origins.

Accept it, live with it, or take it up with the people who use YOUR "rights" to conceal their own shameful and insidious agendas.

Hey, I don't have to accept anything of the sort. FAR more people have died in previous wars specifically to protect my rights, and I'm not about to concede them without a ruckus. And that is what's happening here--people fighting shameful and insidious agendas.
membrane gmoon7 years ago
I could not agree more we should never give up our rights with out a fight.
Right on. Although Dream Dragon is well intentioned, I simply cannot accept the loss of my liberties because of criminal or terrorist actions. Isn't that one of their goals? Like the good man Ben Franklin said, those who give up liberty for security deserve neither (and will probably get neither).
Shadyman7 years ago
Some cameras even save Body Number of the camera.
jerkey7 years ago
Recently I was a victim of a crime, and was talking with the Police Detective about his investigation into the main suspect. He told me that person was ruled out, because he had checked into the location of his cellphone at the time of the crime, and he was not there (or at least his phone wasn't there).

Most people will think you are paranoid if you tell them that "they" are LOGGING the physical location of your phone at all times, Just In Case they need to look at the data later. The three-judge panel whose job it is to authorize warrants for that data elevated their standard of evidence awhile by one level awhile ago, but not much has changed.

And on an even more relevant note, people doubting the government's determination to track people's activities using cameras and printers should look at this:


they really are after you.
oncex jerkey7 years ago
I knew about the yellow dots but it would be nice to find a site with software patches for the printers. But I'm glad I don't live in a place like China.
sedition7 years ago
Lawl. Half of these responses are simply precious: it's like no one has a friggin' clue OR sense of humor. Great Instructable and great book, w1n5t0n.
Yeah... and I was the one who got censured. (I thought I was funny - I was trying anyway.)
Rishnai7 years ago
Seems to me like most people wouldn't have to worry about the federal government targeting them, but there are many good reasons why the average citizen might be picked on by local "law enforcement." So yes, it is a little paranoid (but not necessarily a bad idea) to worry about some CIA contract on your head for taking landscape photographs, but taking a urban landscape photograph and then posting it could easily lead to some pencil-pusher deciding that you've taken one too many photos of the U.S. mint. Further, photographing, accidentally or on purose, police activity is legal. That doesn't mean the cops won't hassle you, railroad you, or try to find an excuse to book you or charge you with something. For example, the people that film protests with Denver CopWatch regularly get harassed, their cameras sometimes confiscated, and ALWAYS asked, told, then ordered (with no actual authority) to shut off their cameras. Therefore, it is reasonable that the Denver police may look up internet postings of photos of their conduct, compare signatures, and then look for a photograph that they can get a charge or citation to stick to, based on that signature. That provides just one more probable cause to arrest a member of Denver CopWatch, a retribution for filming them.
ELF7 years ago
I think that masking the image with noise on all 3 channels might be a very effective way to remove the signature. It shouldn't be photoshop's noise though, because I don't think that's truely random. A quick google gave me this:
Using that to generate a new pattern every time would make it impossible to find out what the signature pattern looks like, because you would never know how much lighter/darker each pixel is, without any reference.
Unfortunately, it only gives 1/0 results, so either total black or total white pixels, which is propably not enough...
Jake Turner7 years ago
Very nice idea, but the only real use I'd see for this would be child porn or photos from area 51.
merseyless7 years ago
also, never print your pictures
and go to the
"Steganographic anti-counterfeiting" part
That only applies to certain Laser printers. :)
If you want to be 100% sure, get a Oki, Samsung, or Certain Xerox Models, because they do not print tracking dots.
Here is a list of printers that do and do not print them:

Also, Inkjet printers do not print tracking dots.
locofocos7 years ago
we know what SOMEBODY has been doing with their camera.....
nature2237 years ago
how about simply just using your camera and not worrying and being paranoid, and doing the:"THE GOVERNMENT IS ABOUT TO GET ME FOR TAKING LANDSCAPE PHOTOGRAPHS" person... or,simply use TWO or THREE different cameras,ebay them MONTHLY,or do a random sell pattern so then every so often,or maybe mayyyybe,oh yeah reality check...sheesh,talk about conspiracy theory's run amuck....dude,do you realize just HOW many pictures are on hosting sites? and why would a generic outside picture make them "home in on you"? just don't take pictures at OBVIOUS Government facilities Ie; SECRET SERVICE/CIA/NSA/USDA/BATF that will also "take pictures of you,because you took pictures of US..first" a little common sense makes alot more sense..good info though,thanks
Ever lived somewhere where simply voicing any negative thoughts about the current regime will get you killed? Obviously not. Just because you live somewhere where you don't have the secret police making people disappear nightly doesn't mean the rest of the world is that friendly.
i'd suggest you'd move then
Tor isn't inherently a better or safer idea, unless you use full end to end encryption independent of Tor. It encrypts within the network, but the exit nodes are not encrypted. Thus the exits are good choke points to sniff at.

@ mightysinetheta -
End to End encryption (e.g. PGP encrypted files or emails, or SSL secure web sessions) on its own might protect the Confidentiality of your uploaded images in transit or storage, but does nothing to preserve your Anonymity from powerful snoopers who have access to either your computer's Communications Traffic Data, or that of the target upload computer i.e. host IP address, time, date, amount of data uploaded, and possibly other web browser environment variables etc.

If, for example, you are uploading pictures of riots being brutally suppressed by the authorities, and yours is one of the few, or perhaps the only, upload session to a particular website, from say, Burma, Tibet or Zimbabwe, then the fact that you have used End to End Encryption will make little difference, as you, your family, friends and associates, are hunted down by the local secret police.

One approach may be to use an encrypted Secure Sockets Layer web server upload session (https:// as used for online credit card and internet banking transactions) in combination with a Tor Hidden Service running on the same machine.
e.g. one of the methods of secure anonymous uploading offered by the wikileaks.org whistleblower website:


This is not that easy or swift to use, and is not infallible either - you have to trust that wikileaks.org or their former Pirate Bay web hosting company in Stockholm, Sweden, has not been infiltrated. Anybody posting banned or controversial images, has to make their own risk calculation, balancing risk of discovery with speed and convenience.

Wikileaks archives of censored images of riots in Tibet
@ WikiLeak and NuclearDog You guys are correct, I didn't mean to say that you shouldn't use Tor or that SSL was better alone for this application. (Should have worded my post better) What I wanted to point out was that Tor has its own issues as well, some of which are remedied by using SSL in conjunction with Tor. I think WikiLeak covered it very well in his post.
Tor is an inherently better idea in a case like this (posting images). The link you posted discusses the ability for an exit node to sniff the traffic coming out, but I'm pretty certain the author of this instructable is more concerned about anonymity (which would still be preserved) than someone getting a hold of some of his traffic (only some, as tor will change exit nodes each request). He's talking about posting the images online anyways, so someone sniffing them and getting a copy is really not a concern at all.
Gjdj37 years ago
very nice. good job. what other cool instructables are you planning?
ahhhhhhhh the government!!!!!
Conexion7 years ago
While this is nicely done... honestly, if you're that concerned about autonomy on the internet, and can't figure these sorts of things out on your own... then you most likely aren't smart enough to stay anonymous on the internet. Oh yeah, try doing a light median filter on each colour channel, then reduce the size. This keeps your quality up while mostly removing your noise signature. If you're really concerned - add light noise above that and redo the median filter.
murfs7 years ago
Hmmmm.... What's about simply, 'show' the screencap the picture and paste and save it again to medium/low jpg? Wouldn't the 'presentation' by the OS simply kill the original noise/ add own ?!
nope, because the "noise" that he's talking about isn't actually some psuedo-spooky gov't hooba-jooba, but static introduced to the image as a side effect of the circitry of the camera. the thing to remember is that every wire in any peice of circitry acts as a mini-anttena, and that when the EM of the surrounding curcitry is introducied, it will have a reproducible effect, creating, in essance, a "signature" that will show up on any image that the camera takes. the way to get around this is to either 1) introduce new static that will mask the original signiature by adding superflois curcitry to the camera (make sure it's drawing off the camera battery) 2) artificially do the same w/ image editing software. it's easier to do the later simply due to the nature of computers. the best way of doing this, of corse, is to reduce the image to the point that any fine detail information is lost in the resizing process. my $.02
I understood that pefectly, But my guess was that the limited coulourdepth of the monitor would 'kill' the variances in the pixels, hence 'noise'. And in a secound step, screencaptureing it and saving this caputere as a medium quality JPG copy would basicly evade all additional information. It would just be a dump from the screencontent! Your basicly going from digital file over presentation, back to digital copy.
CyberBill murfs7 years ago
I dont think you understand how "screen capturing" works. If you just hit the "Print Screen" button on your keyboard or use a screen capture program, there is no 'presentation' as you say. You are literally copying the bits from the frame buffer, so its still a digitally perfect copy.
ax0n CyberBill7 years ago
You could always use a friend's camera and take a picture of your own monitor. ;)
ax0n7 years ago
I found this article highly interesting. Let me share with you a tip I use that might help:

When I'm scanning in images from a print source, I invariably see all the dots used to make up the image. The trick with those is to scan them at the highest optical resolution you can (1200 dpi or so). Zoom in to 100%, then apply gaussian blur. Push the radius just far enough that the dots start to run together. Then under resize image, chop the DPI in half, choose "bicubic sharper" and let it go.

The result is a smooth image, and usually I apply a gentle unsharp mask afterwards to restore some edge contrast.

In this case, you'd probably do well to just leave it a tad blurry. Any radius over 1.5 pixels should be enough to knock out the noise signature after it's resized.

Additionally, there's a very nice Perl script called exiftool that's part of the perl-Image-ExifTool module. It allows all kinds of scriptable exif manipulation, but the most important, is the ability to destroy the exif data completely.

Example: exiftool -all= <filename> (note the space after equals)

The article also mentions digital noise reduction. My Canon DSLR has such an option. (All Canon DSLR's offer this, from the original digital rebel up to the EOS Mark 1 DS) What it does is takes the shot, then takes the same shot with the mirror down and aperture closed to get a "noise print". It then subtracts the second data from the first. That should help slightly, although it may be possible that this creates a "hole" that can be identified also.

As the article says, any set of manipulations repeated will eventually hand someone a pattern should they choose to undertake such an option. (No doubt they will.)

To wit, you should take as much precaution in how you post your image as you do in scrubbing them to prevent them from being used against you.

Tor is a good start, but not a complete solution in and of itself. Consider stealing wireless from somewhere and then using Tor. Additionally, using services hosted outside your country (preferably ones that aren't cooperative with your government) can be a good road block also.

Lastly, remember that the truth will set you free!
I've worked as a professional astronomer, and I know a lot about methods of noise reduction in images. The "evenly illuminated screen" method (called a flat-field) described is useful for taking out some portion of unchanging image noise, but there are other methods that we use as well. The one that's practical for use with regular digital cameras is the "image flat" -- we would take all of the images taken over the course of the night and average them together. The signal averages out given enough image, leaving only the unchanged noise -- you can then subtract it out from your actual images. The software we used doesn't work with JPGs, and it's a lot easier for monochrome images (you'd probably want to work separately with the blue, red, and green channels) but the principle should work just as well. Take a bunch of pictures of something fairly boring -- the sky on a cloudless day, a white piece of paper, whatever -- and average them together to get your flatfield.
I seem to recall a program called Vega that could do dark frames, and that sort of thing.

That could be quite useful, as presumably, subtracting the dark frame should do that. I think it works with jpgs.
I'm not sure where to find a copy, but whilst looking, I found a program called Astromix that looks promising.
Its at http://www.astromix.com/download.htm
Valche7 years ago
While I doubt that I'll ever be in a situation where this will be necessary information, and although it is a bit paranoid, this was a well done instructable and should I ever be in a position to make use of this info i'll send my thanks accordingly :).
Zak7 years ago
Good to realize that this exists and can/should be countered at least in some cases. As for the argument that adding noise will not remove noise that is there: that is true. A 6 megapixel image has 6 million independent pixels. I suppose JPEG removes a lot of dependency between pixels, but let's say that in dark environments there remain half a million 'noise' pixels. If every noise pixes has + or - 4 brightness values, how much noise would you have to add to block identifying a single picture? Probably adding the same nose again (including the filtering that the camera applies to the noise). What is you needed to identify two hundred? Better add the same noise to every picture, or the pictured can be averaged again to reveal the constant noise of the camera. By using the same noise for every picture, it just makes them look to be 'from the same camera', and averaging them will not reveal the underlying noise pattern.
CyberBill7 years ago
I'm surprised nobody has posted up here yet about what I'm about to talk about! I'm an amateur astronomer, I do a lot of astrophotography using regular off-the-shelf digital SLR cameras such as the Canon Digital Rebel XT / 350D. There are three main ways we remove noise and other imperfections from our images. 1) Dark frame subtraction - Take two pictures using the same ISO sensitivity, the same exposure time, etc. One of which has something (lens cap!) in the way so that no light can enter the lens & hit the sensor. The other has the lens cap removed and looks like a regular image. The Dark Frame will look almost entirely black, but will be speckled slightly, which represents the "digital signature" of your camera. Remove (literally subtract the pixel values) of the dark frame from the regular frame, and there you go. We typically take 5 or 10 dark frame images and average them together first (to even out random noise). 2) Flat-Field subtraction: Take a picture of something evenly illuminated to about 50%. Try a white piece of paper or something similar. Dark-frame subtract this image to remove 'constant' noise. (Also, take 5-10 images, and averaging them helps) Then you can use the pixel values in the image to make individual pixels slightly brighter or darker as compared to the average pixel brightness. This will remove dust and lens defects from your camera (particularly vignetting). 3) Image averaging: This isnt particularly applicable because it only removes the truely random noise and, thusly, isnt a 'signature' of your camera. But basically, take a bunch of images, do dark frame subtraction and flat field correction on them, and then average all of those images together. It will remove the random noise in the image. Using #1 and #2, combined with removing the EXIF information from the jpg will make your images untraceable as well as keeping them BEAUTIFUL, without having to introduce more noise into the image. As a note, adding extra noise to the image will most certainly not make it any more difficult to trace an image back to your camera because the original noise is still there, its just behind other noise. Good luck keeping the spies off of your back. :)
Do I smell a photoshop tutorial on how you do it? That would be awesome, even if it is something very easy to do. I dont know how to do it, but I would be interested in knowing how its done.
cactuar127 years ago
Finally, now I can post CP and not have to worry about the Feds :P
nope if you post CP the feds have enough reason to check with tor or any proxy you use who you are even if you use 7 proxies you will get v& if you do it often enough
I know you're just joking, but that's just sick and messed up.
Digitally scanning a film image is really just another form of digital photography, so you should really also apply Steps 1 to 5 to your scanned images as well.

Scanners also have their characteristic individual electronic, optical and mechanical imperfections and abberations within their manufacturing tolerances, and they also pick up extra potentially identifying charateristic wear and tear with use.

If as Step 7 suggests, you make use of someone else's scanner, then you might not be able to erase the original copies of the temporary files which are written to the hard disk of the personal computer running the scanning software, or which are sometimes built into the the larger types of office or commercial scanning / photocopying systems (or even Johnny Mnemonic style high resolution fax machine memory buffers).

You will also have probably have left your fingerprints and DNA samples on the equipment.

Photoshop and other image manipulation software tends to preserve most of the identifying EXIF metadata, which can include potentially traceable digital camera serial numbers, Adobe software GUIDs etc.even if you apply filter effects (some of which are reversible) or resize the image, unless you Save As to a different file name from the original.

Most digital cameras (and mobile phone cameras) also embed a small thumbnail image which they use when they display a "photo album" view of the various images you have taken, on the camera / phone display , usually to aid image selection for deletion or transfer etc.

This is not amended when the main image is manipulated in Photoshop etc, and may betray or give clues about identifying numbers or human faces etc. which you are trying to censor or redact.

Phil Harvey's multi-platform ExifTool perl libary and stand alone Windows executable software, allows you to view, amend , add or remove such metadata (and extract embedded thumbnail images), from a wide variety of image and document file types.


digitally scanned images don't need to be edited if you don't use your own scanner you may leave traces to the copy store, but never to your own home
Scanning the image is really the only way to be sure; even noise added by cropping and recompressing won't disguise everything, according to Fridrich's work.
romulusnr7 years ago
Couldn't you take a picture of some evenly lit neutral scene, or a series of such, then use the variations in such images as a negative overlay? In theory this would counteract the noise by applying "negative noise". Just like an audio program's noise reduction: sample a segment of dead air, and apply the resulting noise inverted onto the rest of the audio. (well, ok, it's not really like that.)
m1k3y (author)  romulusnr7 years ago
You could average together many images as long as nothing changes to reduce noise and get rid of the noise signature. This isn't really practical since its hard to take 10 identical pictures of anything that could possibly involve movement. To find the signature itself involves complex noise filtering algorithms. Basically you are carefully removing the noise from a large series of images and then statistically comparing the noise for similarities. Once you found your signature, you then need to verify it against a series of new images from the camera and images taken from other camera. If you can consistently pick which ones came from your camera, then you have done it.
The link you provided showed a bio of a woman who specialized in analyzing photographs (such as photoshopped forgeries) and worked for the military. If you drew you inspiration to get this worked up off that one paragraph, you need help. As far as the instructable: 1) Images posted on the web are resized 99.9% of the time. This makes pixel level artifacts nearly impossible to find because the web version has lost a majority of it's data and each displayed pixel is recomputed from an average of several pixels. A large photo on the web would be around 800x600 pixels. That is less than a half-megapixel. 2) Dirt specks on the lens do not create tiny artifacts. If they are visable at all they are scattered over a large area and are extremely blurred. The reason is that they are incredibly out of focus. Furthermore, they will not stay in the same place on the sensor (like dust on the sensor will) because the acts of focusing and zooming will change the characteristics of the size and position as well as relative blur. If you're really concerned about small bit of dust it may further unnerve you that they are on every internal element of the lens which is not cleanable. 3) Dark frame subtraction (what astronomers and DSLRs on long exposures use) is to remove fixed pattern environmental noise. That is why it is done for every shot - because it is a unique interaction between the camera and a particular location/temp/humidity/etc at that moment. It would not be reproducible for forensics. What you would be trying to eliminate would be fixed pattern noise caused by the camera itself - something that is not nearly as prominent in modern digitals as it once was. Basically the analog amplifier noise and read-out noise. To target such noise, you would need a program that could compare a decent sample size (over 50) of images shot in all different environments to create a noise profile for the camera. The profile would have to be dynamic though, as changes in ISO and duration of exposure would influence the strength and characteristics of the patterned noise. Then you would have to apply the negative of the noise image to everything you shot. Considering how dynamic the noise can be from a single camera, I'm doubtful of identifying a camera beyond reasonable doubt even from raw files. From downsampled web images, there is no way.
Here's some more links for you, if you really want to learn more...

Determining Digital Image Origin Using Sensor Imperfections (pdf)
Digital Imaging Sensor Identification (Further Study) (pdf)
Identifying Common Source Digital Camera From Image Pairs (pdf)
Camera Identification from Scaled and Cropped Images (pdf)
Camera Identification from Printed Images (pdf)
Binghamton University research links digital images and cameras
Digital Cameras Have Unique Fingerprints

The important thing to keep in mind is that the noise pattern is "spread spectrum". I.e. it's not just uncorrelated pixel noise, and you cannot just erase it by scaling, cropping, compressing, filtering or adding noise (although each of those would make detection harder).
m1k3y (author)  Forgot_Password7 years ago
I understand your concerns, but there are environmental and hardware situations that can create static identifying marks. All you would need to do to test it is put a camera in a completely dark space, take lots of pictures and then algorithmically increase the contrast. Watch the images as a stop-motion movie. There will be pixel groups that stay static. This isn't a noise signature, but is a unique artifact that can be used for identification.

I disagree with your complaints about resizing since the signature is usually determined through statistical algorithms. Resizing will make things harder, but not impossible. Cropping the image will make it harder but most images aren't cropped.

If you got past her picture you would have found pdfs of all of these research papers on the subject and if you dig a little further you might find some helpful C++ code examples.
I don't get the point of doing that. What kind of 'legal' pictures could make someone paranoid like that ?
m1k3y (author)  chooseausername7 years ago
If you live in China, North Korea or even in one of those small former USSR countries, "legal" is open to interpretation. Or if you got a picture of someone really important doing something very naughty, then you might not want to be connected to it. This is also helpful when someone in the press has a source they don't want to reveal.
ll.137 years ago
Paranoid! xD -For EXIF data removal I like "JHead" :)
Very nice explanation. Thank you.
buttersnake7 years ago
I have never really needed to try to hide who is taking the pictures that i take, but if someone is sitting in a room lined with aluminum foil and eating canned beans everyday for fear of going outside because of conspiracy theories they have in they're head then i guess this would be great.