Computers, Court, and Fingerprints

Degrees writes "Should Law Enforcement be allowed to Photoshop fingerprints? That is the question posed in this article in the South Florida Sun-Sentinel. The suspect is charged with murder, and the evidence was circumstantial before the fingerprint enhancment. At the end, the crime scene investigators say they want encrypted cameras. The implication is they want DRM-enabled digital cameras with software for full audit-trail capability. Would that make the Photoshoping more credible? Would DRM cameras be a good thing for Joe Citizen?"

Comments

"Enhanced" evidence

[russotto]

They've tried this with audio before, notably in the Waco cases. The court rejected it then. Hopefully they will keep rejecting it. Such digital enhancement might be useful for getting leads, but the result isn't evidence; it's just a computer-assisted guess.

Re:"Enhanced" evidence

[sacherjj]

Can't you look at the enhanced and unenhanced and verify that it is the same print? To me, enhancement can be considered part of the collecting, sort of like a filter that may exist in an audio recorder. If new ridges were not drawn, is this truly "doctoring" the print? I agree that it is a slippery slope and the determination has to be the level of enhancement. Can it still be considered a cleaned up version of the original, or did the enhancement add new ridges, etc. to the print?

Re:"Enhanced" evidence

[Jason Earl]

Spot on! I don't mind if the police use this kind of thing for leads, but this shouldn't be considered "hard" evidence.

Re:"Enhanced" evidence

[Planesdragon]

Hmm, you can't really use it to get leads either. Any evidence that comes from inadmissable evidence is inadmissable itself.

I believe that you're wrong.

If you're given an anonymous tip, or someone takes a lie dectector test, you can't use these as evidence--but you CAN use them to get evidence.

Anything that comes from ILLEGAL police procedure is tainted. Not just inadmissable evidence.

Re:"Enhanced" evidence

[CaseyB]

Such digital enhancement might be useful for getting leads, but the result isn't evidence; it's just a computer-assisted guess.

Applying, say, a contrast filter to a digital image to bring out details is no different from the subjective treatment that a conventional photograph gets when developed in a darkroom.

I imagine that the various tests that forensic scientists perform are rigorously standardized. There's no reason that digital processes couldn't be similarly regulated. I supposed what is called for is the certification of "official" digital filters, that are analyzed and confirmed to manipulate the image in an "unbiased" fashion.

Re:"Enhanced" evidence

[antis0c]

Yeah, like I said it's a little more complicated than my example. It's really up to the judge what is an isn't admissable.

Re:"Enhanced" evidence

[aero6dof]

I don't think your argument applies to this issue. Why would enhancing a fingerprint change a legally obtained fingerprint to one that was illegal?

Lets say police legally obtained a fingerprint and DNA evidence on a rape case. If an enhanced fingerprint leads them to a known address of the individual then I would think that you have a perfectly reasonable basis to demand a DNA sample of the suspect to see if they compare.

If the original un-enhanced evidence was illegally obtained the enhancement isn't going to change the state of it the other way either.

Re:"Enhanced" evidence

[GlassHeart]

the result isn't evidence; it's just a computer-assisted guess.

Nonsense. I've seen with my own eyes how a fuzzy security camera or satellite photograph clearly show faces or license plate numbers using sophisticated software in several major Hollywood movies.

Re:"Enhanced" evidence

[terrymr]

I imagine that the various tests that forensic scientists perform are rigorously standardized. There's no reason that digital processes couldn't be similarly regulated. I supposed what is called for is the certification of "official" digital filters, that are analyzed and confirmed to manipulate the image in an "unbiased" fashion.

Yes they are - but on the other hand there have been many miscarriages of justice because the jury thought the forensic scientist was giving them hard facts rather than just his opinion. Circumstatial evidence of any kind is open to more than one explanation.

Re:"Enhanced" evidence

[Alsee]

but the result isn't evidence; it's just a computer-assisted guess.

It seems real simple to me. Give someone the evidence to enhance all they want in absolutely any manner they like. Just don't give the person doing it a copy of the suspect's fingerprint, image, voiceprint, whatever.

If the result matches the suspect and does not match anybody else then it sounds like solid evidence to me. There is no way you can photoshop someone's fingerprint into an image if you don't know what his fingerprint looks like.

Even better make it a seperate person who checks for a match. Even better give that seperate person a dozen random fingerprints and don't tell him which one is the suspect's. If he says there is a definite match AND he says it to the print that happens to be the suspect then you have a pretty damn bulletproof system. It would be pretty serious event if the expert ever reported a "definite match" to one of the extra random prints he's given.

At that point I don't care if the image was "enhanced" by a chimpanzee twiddling an etch-a-sketch.

-

Re:"Enhanced" evidence

[1u3hr]

Hmm, you can't really use it to get leads either. Any evidence that comes from inadmissable evidence is inadmissable itself.

No. Any evidence that comes from illegally obtained evidence poisons anything derived from that. Much of the evidence gathered by investigators would not be admissable in court, but gives a clue as to where to look for admissible evidence. Even asking a psychic for a clue would be fine, if it lead to finding, say, the murder weapon, though of course the psychic's testimony would not be evidence at all.

(No, I don't believe in psychics.)

Re:"Enhanced" evidence

[Reziac]

So what about *printing* the digital image, then applying standard photographic techniques to bring out details? Yeah, I know there's a problem with DPI compared to film, but at least it would be a sort of proof of concept. As you say, there's no reason that the digital tools can't be standardized.

The real question, IMO, is the audit trail and whether someone might have dropped a few pixels in there manually while they were applying the filter. The solution being to blackbox the standarized enhancement process, tho you've still got to keep track of who has access to the file before and after processing.

Re:"Enhanced" evidence

[WNight]

This is like saying that records are exact duplicates of the original music because they're analog. Even if the recording process was perfect, the materials aren't and can't hold exactly the shape intended.

The photograph is limited by the (fairly large) size of the light-sensitive grains in the film.

I'd trust digital processes more because you can keep all in-between steps and use only documented sharpening routines, etc. Both sides could go back to the same digitally signed original and follow the steps themselves. With optical processes you trade your distrust of digital for completely untraceable analog steps. Who knows what was done to the picture? Nobody can ever repeat an analog step exactly the second time so you can't ever get the same result from the negative. Was it because it was intentionally added, or because there was just a small mistake during one of the processes?

Using digital is fine if you scan the negative at a high enough resolution to capture the film grain (and thus be assured of capturing all detail) or if you shoot at a resolution that provides more detail on resolution charts than the analog film which would be used.

Re:"Enhanced" evidence

[Fat Casper]

I would think that you have a perfectly reasonable basis to demand a DNA sample of the suspect to see if they compare.

How do police demand something? They get a court order. You're not going to get a judge to grant an order based on evidence that he won't allow to be presented in his court. The inadmissability of evidence doesn't change based on what you want the court to let you do.

Of course, it is a legitimate lead. They can interview the hell out of anyone willing to be interviewed. Until the cops come up with something real, however, you got nothin'!

NO WAY

[QuantumRiff]

I don't want them tracing those nude pictures of my 17 year old girlfriend back to me... I prefer to just put them out on gnutella for all to see anonymously! ;)

THis does absolutely nothing

[Anonymous Butthead]

Wait... so they can crack the encryption?
Nothing is flawless, any form of encryption can be cracked.... all you need is time.

It would give the community a false sense of security. Just becasue id has DRm doesn't mean anything. Evidence should not be tampered with.... PERIOD

Re:THis does absolutely nothing

[outlier]

I think what they're actually talking about is digitally signing the original or computing some type of checksum to maintain an auditable chain of custody and to track versions of the data.

So, ideally, an officer takes a photo and it is automatically digitally signed in some form of read only storage. The image and the signature are then transferred to an electronic "vault." Any 'enhancements' would also be signed and stored in the vault. When the case goes to court, the defense is given access to all versions of the picture, and all the images are matched with their signatures to ensure that they haven't been tampered with. This way, the defense can have their own experts evaluate the 'enhancements' that the police made.

In this scenario, you never deal with concerns that encrypted images may be decrypted. You have to have confidence in the vault -- I'm guessing that a physically secured, tamper evident device with easily auditable features could be implemented (e.g, in the same manner as the FBI carnivore machines are secured at ISP sites).

Re:THis does absolutely nothing

[Reality Master 101]

Wrong. One time pads cannot be cracked. It is mathematically impossible.

After that, we can stop playing word games. When someone says uncrackable, they mean uncrackable by any practical method. Practicality is the only thing that matters in the discussion of encryption.

Re:THis does absolutely nothing

[Twirlip of the Mists]

What? That's the dumbest comment I've read all day. Let's define "crack" in this context: to decrypt without access to the necessary key(s). A message encrypted with a one-time pad is decrypted with the same one-time pad by the recipient. It is not possible to decrypt the message without the one-time pad except through simple brute force: i.e., trying all possible combinations of transforms over the length of the message. This method will, of course, produce a very large number of readable messages, only one of which is the correct one. So yes, "cracking" a message encrypted with a one-time pad is, for all purposes, impossible.

Re:THis does absolutely nothing

[Reality Master 101]

A one time pad is a key made up of a bunch of random numbers exactly the same length as the message. So one side encrypts with the pad, while the other side decrypts with the same pad. It is mathematically impossible to crack the code without the pad.

Re:THis does absolutely nothing

[Reality Master 101]

Here you are, completely unbreakable encription.

You are correct.

One-time pads require that your key is as long as the message. Hello?

And the problem with that is... ?

Encryption is encoding a message in a secret way such that it cannot be read by unintended recipients. If I'm sending news from deep within my oppressive government to share with the world to my partner in a free country, and I use a one-time pad that we exchanged in the past, is that supposed to not qualify as encryption?

Re:THis does absolutely nothing

[Reality Master 101]

Sure they can; just get your hands on the pad.

Stealing the key is not cracking the code.

But how do you exchange the pads securely?

Here's a CD with the pad. Don't lose it. Not every communication has to be electronic.

Re:THis does absolutely nothing

[Tassach]

Your Adam/Eve scenerio is a CODE, not a CYPHER (which is what we typically mean when we say "Encryption"). Codes have a limited vocabulary, bounded by a pre-defined codebook (which you trivially illustrated.) A cypher obfuscates any arbitrary message algorithmically via a key. A one-time pad is merely a cypher with a key length equal to the message length.

Codes are most definately breakable. If the code book is compromised, it's broken. If you use the same codewords over time, it is subject to analysis. If your codewords map 1:1 to words a natural language, it's breakable. If your actions can be correlated to the message, it can be broken. And so forth. A code is only 100% secure if the codebooks remain secret and it is only used once. "Climb Mount Nitaka" only works once as a code for "Attack Base PearlHarbor". Having used this, using "Climb Mount Fuji" as code for "Attack Base Midway" is really stupid; the opposition might not know exactly WHERE you are going to strike, but they know an attack is emminent. If you sent a previous message "Visit Mount Fuji" to mean "Photograph Base Midway", and I correlated that message with a recon flight over midway, now I know exactly where the attack is coming.

Re:THis does absolutely nothing

[kesuki]

There is nothing special about 180-210MB (21-24 minute) 80 MM CD-r. They're industry standard. you can buy a 10 pack at wal-mart.
But yes, DRM isn't needed, all that they need is to make absolutely sure that when 'images' are enhanced that the original remains available and untouched.
Manipulating pictures with filters should never ever be the sole piece of evidence in a case. It should be just one of many pieces that leave no room for doubt.

The glove didn't fit

[CrazyDwarf]

I've always wondered about cases where digital evidence was envolved. We have no way of knowing if the files are tampered with or otherwise altered, and I really doubt they'd let us compare (in this case with actual fingerprints.)
I think that guy that was on trial recently for the disappearance of that girl didn't surf for child porn, the cops did... then changed the dates on the files to cover their own butts.

Re:The glove didn't fit

[JoeBuck]

This is nothing new; many Photoshop-style transformations are easy to do with traditional photographic print methods (dodge and burn, editing out objects you don't want to be there, blurring, cropping, etc).

Re:The glove didn't fit

[evilviper]

I've always wondered about cases where digital evidence was envolved. We have no way of knowing if the files are tampered with or otherwise altered

*Ahem*. So you wouldn't believe what you'd see on the tape if I captured video of bigfoot on a digital camera, but if it's captured on film it must be real?

And you'd put faith in a 35MM film photo of nessie, but not one taken on a digital camera?

Just because most people can do basic manipulation doesn't mean that anyone can do a convincing forgery that would convince the experts.

I think that guy that was on trial recently for the disappearance of that girl didn't surf for child porn, the cops did... then changed the dates on the files to cover their own butts.

Well, all it takes is a computer forenics expert to analyze the hard drive, and he can definitively tell you if that was the case.

Forge a signature, forge a timestamp, either may pass at first glance, but won't pass close scrutiny by an expert. It's the case with digital, as well as analog. In fact, computers can more effeciently analyze digital sources, so digital may be even harder to convincingly forge, than analog.

I saw this on tv

[LennyDotCom]

I saw this on the discovery channnel I think they showed how all the cop did was enhance the image with photoshop. All he did was apply a custum filter to enhance the image he didn't add anything to it or change it just brought out what was there by filtering out the background. I was very obvious if you saw the show that it should be perfectly legal .

Re:I saw this on tv

[LennyDotCom]

Yes I knew somebody would mention that but his filter was just to neutralize the background. In the show he was getting a fingureprint off a sheet or some kind of fabric all the filter did was match the fabric pattern and remove it leaving just the fingure print. If he added to it it would be a different story but deleting what you don't want and leaving what is left isn't tampering with the accual fingerprint

Re:I saw this on tv

[LennyDotCom]

accually I can't rember the details but another poster does a better job of explaining it the technic is really just dodge and burn where they only lighten or darken the image so they lighten up the background and darken the print If you saw the show yopu would agree that there was no doctoring the print

Re:I saw this on tv

[Ibag]

I don't think that the question is whether this should be legal. The question is if evidence gathered in this way should be admissable in court. While this isn't the same as the cops taking evidence from your house without a warrent, you do have to worry about the accuracy of the technique and whether it should be allowed in court. They mention in the article about going in and digitally removing "interference" like weaves of fibre. Who is to say that the removal technique is good enough to recover the fingerprint exactly? What if the removal process adds/subtracts features from the fingerprint itself to the point that it appears to be a match but might not be?

It seems to be a useful technique for gathering evidence to find a suspect, but I'm not sure that I'd want it to be key evidence for a conviction.

Re:I saw this on tv

[dnoyeb]

I saw the show, and no it should not be perfectly legal. The resultant image will be different based upon how the user conducts his technique. That means its more art than science.

Its not prosecution or defence quality evidence. Only investigative.

Re:I saw this on tv

[WowTIP]

Yes, but if the filter was designed to alter the fingerprint (which imo is hard to impossible to do), the result will be the same in the courtroom as in the cops lab. They couldn't use funky Photoshop tools at it though.

But the question remains. If the cops wanted to frame the suspect by altering the evidence, why not plant an object with the suspects fingerprint at the scene to begin with. Or replace one piece of evidence later. Much easier than manipulating existing fingerprints in graphics software.

Re:I saw this on tv

[IceDiver]

I saw this on the discovery channnel I think they showed how all the cop did was enhance the image with photoshop. All he did was apply a custum filter to enhance the image he didn't add anything to it or change it just brought out what was there by filtering out the background. I was very obvious if you saw the show that it should be perfectly legal.

So it should, as an investigative technique. But to be admissable in court it needs to be shown that no other changes were made. To do this would require an in court demonstration. Something like:

"Here's the original fingerprint. We will now scan the fingerprint."

"As you can see (shows jury), the scan matches the original."

"I will now open and install this original copy of Photoshop. (Installs). I have a representative from Adobe here to verify that this is, in fact, an unaltered copy of Photoshop. (Adobe rep verifies installation.)"

"Now you may observe as I use the following standard technique to clarify the scanned image. (Uses big screen so entire jury can observe.) As you can see, the enhanced image matches the accused's fingerprint."

Using this method, the fingerprint can be entered into evidence, and the jury can be reasonably certain that no underhanded alterations were made during the process. However, the defense can also attack the evidence in a number of ways: arranging expert witnesses to attack the reliability of the method, demonstrating an alternative clarification technique that provides a different but equally convincing result, and so on. In this way, the prosecution can present their real evidence, but the jury is given proper opportunity to weigh its reliability and relevance.

IceDiver.
Politically Incorrect - and Proud!

Re:I saw this on tv

[5KVGhost]

That's right. The kind of enhancement they're talking about is apparently just a bit of unsharp masking and contrast adjustment. The same sort of stuff many digital photographers do routinely. And if they were doing it in a darkroom with a film negative no one would even raise the issue.

Of course there is the opportunity for abuse or falsification, but that's the case with any kind of evidence. Unless the defense is claiming some kind of deliberate tampering with the fingerprints then they're just taking advantage of people's ignorance.

Re:I saw this on tv

[Cyno]

We should just assume you're guilty until proven innocent. That way we don't have to spend all this money on image filtering software OR put up with criminals constantly professing their innocense. God Bless America!

Re:I saw this on tv

[terrymr]

Even a side by side comparison in court of the original and the enhanced print would convice most people however in this case the original print is missing in action which is probably why the defense is raising an objection at this point.

There seems to be a lack of credibility where you claim you found a print at the scene but then can't / won't produce the item bearing that print in court.

Re:I saw this on tv

[drinkypoo]

Right, and in the same vein, this is why we need DRM cameras for law enforcement use. It would be best if all cameras had it for the purposes of random people taking admissable photos but it makes no sense for a digital photo to be admissable evidence at this time, so it would be a neat feature.

"Here is the original DRM-protected photograph..."

Re:I saw this on tv

[ebyrob]

I saw the show, and no it should not be perfectly legal. The resultant image will be different based upon how the user conducts his technique. That means its more art than science.

So engineering is an art form eh? Assign 20 developers to write a large program with a given set of requirements. You won't find two solutions alike, unless someone cheated. Science is all about reproducibility, not "identical results". Even an undergraduate student of science should know that no two expirements will ever produce the same raw data... The goal of science is to "understand" or "accurately predict" those results with the "simplest" theories possible in a repeatable fashion.

A better measure of reproducibility in this case should be who is matched and with what degree of certainty after applying the technique on a given set of "test" prints. If different technicians generally end up with the same matches, then it is "good science" regardless if the intermediate steps differ.

Re:I saw this on tv

[outsider007]

of course this will only work if this slashdotter had the means and motive to kill Nicole Brown.

Re:I saw this on tv

[Reziac]

Well, that depends. The customized filter could be rigged to only enhance certain traits that are known to be part of, but NOT EXCLUSIVE TO, say, fingerprint ridges. So you run the image thru the filter and suddenly, because the existing data fell into the trigger range, pixels get enhanced that shouldn't have been.

Having amused myself by mixing filters and data in all sorts of weird ways, I can tell you that often the results are not simply enhancement. Data disappears, artifacts appear, either of which could create spurious fingerprint ridges. So aside from the possibility of simple pixel painting, it is open to abuse.

I don't think the principle is bad, but rigorous standards would have to be developed and applied, including a complete audit trail (and incremental files) for each image.

Re:I saw this on tv

[dnoyeb]

Yes, but in engineering the solution is not the answer. The solution produces the answer. So their could be many solutions that arrive at the same answer.

Here this tool produces different answers. The results should be beyond a resonable doubt (this is criminal law), and I don't think they will be.

woah!

[RyLaN]

I could to that!! Get a print of my enemy, photoshop it and presto, we've got a conviction! Do the judges have any idea what can be done with Photoshop in the hands of someone who has used it before?

Re:woah!

[carpe_noctem]

As far as the general public goes, they generally tend to over-dramatize the amount of "evil" that technology is responsible for. That being said, I'm sure that judges and legal panel have been made *well aware* about the true nature and uses of this technology before it is able to be used in a court of law.

That's not to say that the jury still won't have the ability to throw such evidence out. They are members of the "general public" still, and as far as lawyers go, the dumber the jury is, the better. ;)

DRM is like the SUV

[aengblom]

DRM is like an SUV... it isn't by definition a bad thing. Only current implementations.

DRM is bad because it causes problems with fair use and long term archiving.

SUV's are bad because people use them for the wrong tasks (people moving) and manufacturers prefer huge profit margins to efficient vehicles.

May they have certain (albeit limited) acceptable uses? Of course!

This is an exciting idea

[L. VeGas]

Photoshop the fingerprints so the ridges and whorls are real big.

Like double D's or so.

Sure...

[Grip3n]

...just as long as they don't go effect happy and start making lens flares and drop shadows and start saying "l00k 47 my 31337 gr4ph1x" during court room proceedings.

Re:Sure...

[medscaper]

and start saying "l00k 47 my 31337 gr4ph1x" during court room proceedings.

Not to be picky...bet...erm...how does one say this without sounding like a nerd in court?

I mean, "But, your Honor, l00k 47 my 31337 gr4ph1x!" writes sorta cool, but, "Your Honor, look at my elite graphics!" just sounds...well...

Maybe, "Your honor, please have entered into the record the following words, in this order, with...Oh, never mind. Can I sit at that thing for a sec?"

(Or, if you're an SNL fan, just this : "MOVE.")

Enhanced...

[Wyatt Earp]

No way.

Remeber, LAPD *may* have tried that with OJ, Time "enhanced" OJ's picture to the tune of a big dollar lawsuit.

Someone else mentioned Waco, yea, "enhanced" evidence is bullshit evidence.

Someone can "enhance" anything, even some yokel atop WTC with a 757 in the background.

Secure digital cameras, photoshoping fingerprints, no way.

Even with a "secure" digital camera, there will be wiggle-room to screw with pixels.

This is NOT DRM

[ChaosDiscord]

DRM is about taking options away from users. This is about providing users with a new option: a strong audit trail. You can make a copy of the image using non-auditing software, but that copy of the image would lose it's "seal of approval." The original would remain valid. The end result: cops can make any copies and image manipulations they want that may help them solve a case. But in court they'll only be able to present images with the valid audit trail, ensuring that the image was never mishandled and clearly showing what manipulations were done to it. It sounds like a great idea and I strongly support this option for users. (I am suspicious that it may not be possible... but I'm happy to let people try.)

Re:This is NOT DRM

[Reality Master 101]

DRM is about taking options away from users.

Uh, no. This is an extremely popular misconception by some people, and an extremely popular knowing lie by other people.

DRM is about preserving the rights of content creators. Period.

Now, unfortunately, taking options away from users is a side-effect of most of the DRM schemes out there. But that is a side-effect, not a first effect. People advocating DRM are not evil boogeymen who derive pleasure from your pain like some music industry vampire. They care about preserving their rights in the face of rampant, out of control copyright violations.

Put it this way: if DRM existed that preserved your fair-use rights while taking away your non-right to mass distribute copyrighted material, they would fine with it.

Don't get me wrong. I don't particularly want my fair-use rights watered down, buy lying about these people's motivations just makes everyone looked like thieves in the eyes of lawmakers, and ultimately hurts the cause.

Re:This is NOT DRM

[Twirlip of the Mists]

You're more right than ChaosDiscord was, but you're not 100% right either.

DRM is about managing rights, both those of the user and those of the creator. "Managing" is a very fuzzy word. It can mean anything from enforcing access controls with encryption and digital signatures all the way down to simply providing a way to store information about rights electronically.

A system that stores rights information-- who is allowed to use the media, and how, and when-- in a database with a link to the content file itself is a DRM system. It doesn't prevent anybody from doing anything; it doesn't force anybody to do anything. But it does provide users with a facility to help them manage rights and clearances. My former employer made and sold systems like this, calling them "DRM systems" the whole time, and nobody was ever confused about what was meant by that.

DRM is a very generic term for an entire class of technology products, like "spreadsheet" or "browser." Saying-- or even implying-- that DRM is inherently either good or bad is just about as meaningful as saying that FTP is inherently good or bad.

Re:This is NOT DRM

[dvdeug]

taking options away from users is a side-effect of most of the DRM schemes out there

It's not a side effect - it's the main effect. The point of most DRM is to prevent the copying of material; whether or not that copying is illegal, it's still an option of the user.

Re:This is NOT DRM

[HiThere]

You may define DRM by it's purported purpose. I define it by it's obvious effects.

I can't know the real purposes of something that is being proposed. I can only know what effects I can reasonably predict it will have. Thus, that's what my definition is based on.

The one you are proposing is based on assuming that what a pr guy tells you is true. I find that at best dubious. At best.

Re:This is NOT DRM

[Reality Master 101]

I can't know the real purposes of something that is being proposed. I can only know what effects I can reasonably predict it will have. Thus, that's what my definition is based on.

Well then, don't be surprised when the opinions of you and those like you are ignored as ignorant by the people who make the laws. Anyway who walks up to a lawmaker and says "The purpose of DRM is to take away my rights!" is going to be ignored.

Just for the record, the proper statement is "The music industry is well within their rights to try and prevent rampant copyright violations. But I believe that DRM as currently implemented infringes too heavily on the fair-use rights of consumers. The solution is for anonymous distribution to be illegal (which it already is), perhaps even with criminal implications in extreme cases. Most cases should probably be civil."

Re:This is NOT DRM

[JoeBuck]

DRM is not only about preserving rights of content creators, because it enables the creation of additional rights for content creators that did not previously exist. Copyright law does not give a content creator the right to control private use of a work, only copying or the distribution of derivative works. But DRM lets the creator control use.

Re:This is NOT DRM

[rhavyn]

When big time television producers (or whatever he was) tell me that not watching commercials is breaking the implicit contract with the TV broadcaster but maybe it's "ok to get up and go to the bathroom during a commercial" I'm not going to believe that they are looking to do anything but make me pay for any and every use of the book/music/movie/tv show. Heck, to prove your point, show me one instance of the DMCA (because that was just to protect their copyrighted digital works) being used in a real copyright infrigement suit, not in a "they're using our stuff in a way we don't want them to" suit.

Re:This is NOT DRM

[ChaosDiscord]

DRM is about taking options away from users.

DRM is about preserving the rights of content creators. Period.

If you assume that DRM's goal is to use technology to enforce copyright law, it's still about removing options from citizens. A key element of copyright law is removing options from citizens. The goal is to encourage the creation of creative works, but the actual action is to remove options. I no longer have the option to distribute copies of other people's works. In the absence of copyright law, I would still have the option. (Mind you, this restriction isn't necessarily bad. I also lack the (legal) option to deceive other people in financial matters, to take money from someone by force, or knowingly put other people in danger. I'm perfectly fine with society removing options for the larger good. I support copyright for this reason.) DRM fundamentally removes options. The options it removes may be illegal, but they do remain options.

However, let's revisit the assumption that DRM is tied to copyright. DRM certainly can have a place in situations where copyright is irrelevant. As an employee, my ability to redistribute certain internal documents may be limited by an NDA or other employement agreement (effectively protecting the document with trade secret law instead of copyright). DRM technology might be employed to ensure that I obey the agreement. If I were to get a job requiring security clearance, DRM might work to ensure that the electronic documents I handled remained secret or top secret. The goal is explicitly to remove options (illegal options, but still options).

All of this is why I chose "options" instead of "rights" or "fair use".

Put it this way: if DRM existed that preserved your fair-use rights while taking away your non-right to mass distribute copyrighted material, they would fine with it.

Clearly untrue. Certainly some publishers would be fine with it. Perhaps even most publishers would be fine with such a system. But at least a handful are interested in using DRM to try and restrict fair use to increase profits. Why are various copy restriction flags being enabled on public domain works (here's one reference)? When you're crafting law or code, you have to consider those who will abuse the spirit of the law/code while strictly following the letter of the law/code. Why is the movie industry fighting to ensure that when Casablanca enters the public domain (2037ish), I'll be allowed to copy the movie, but not allowed the tools necessary to make that copy?

Re:This is NOT DRM

[Reality Master 101]

When big time television producers (or whatever he was) tell me that not watching commercials is breaking the implicit contract with the TV broadcaster

And exactly why is that wrong? Who do you think pays for the production of the show?

I don't know, maybe I'm weird, but I actually appreciate the advertisers paying for the shows that I watch. I must just be a corporate tool.

Re:This is NOT DRM

[Frank T. Lofaro Jr.]

MPAA doesn't want you to skip commercials on a DVD, even though that is fair use. MPAA doesn't want you even making an excerpt from a movie for a classroom exercise, even though that is fair use.

Re:This is NOT DRM

[dvdeug]

if DRM existed that preserved your fair-use rights while taking away your non-right to mass distribute copyrighted material, they would fine with it.

I have a hard time imagining an instance where having the computer read it aloud, wouldn't be fair use. (You're not only going to do a public performance by reading a book, but you're going to do it through some crappy speech synth?) But Adobe eBook Reader has an option not to let you read it aloud. Since they currently restrict free use rights, it's obvious that the intent behind DRM extends to more than just copyright violations.

Re:This is NOT DRM

[rhavyn]

Sorry, but if you think we should be sending people to jail for channel surfing during commercials then I really question the truth in your statement that you're for fair use.

What is wrong is advertising is an expense whereby you attempt to attract interest in your product or create brand recognition. If you have no interest in the ads being shown to you, that isn't your problem. By extension to your argument, people who don't read billboards on their drive to work or who show up late to a movie and miss the premovie adds should be thrown in jail too. In reality it just means that the ads are not effective. I dont know where this attitude that consumers are doing something wrong by not watching adds and buying products came from.

Hope you don't use any pop up add blockers.

Re:This is NOT DRM

[Reality Master 101]

First of all, what the hell are you talking about? Where did jail come into the discussion?

Second of all, billboards don't give back value, TV shows do.

Re:This is NOT DRM

[Reality Master 101]

And yet, "Alice in Wonderland" was placed under Digital Rights Management, so that Adobe could control it's distribution.

You're not being logical here. Yes, AIW is in the public domain, but that doesn't mean you can't copyright particular expressions of it. For example, the words of the bible are public domain, but you can't just photocopy a particular publisher's version of it and hand it out. You could, however, scan the words and distribute that.

Just because Adobe chooses to control a particular version of AIW doesn't mean it ceases to be in the public domain. Just go out and get another copy of it if you want to.

Re:This is NOT DRM

[bmetzler]

MPAA doesn't want you to skip commercials on a DVD, even though that is fair use.

It's fair use to skip a commercial? That's a new one. Usually fair use deals with using something, not not using it.

-Brent

Re:This is NOT DRM

[HiThere]

I'm not surprised that they have ignored my complaints. This despite the fact that when writing to them I have observed courtesy and did not assert my definitions as truth.

In fact, in the replies to my letter I was basically lied to. My "representative" (small r, not my Representative) had already comitted herself to support for a law (since dropped), and was reported to be vigorously campaigning for it when she sent me a reply indicating, though not stating, that she would take my opinion into consideration while considering her vote. Turned out she'd taken campaign donations from sponsors of the bill. (Which was tabled, perhaps dropped, no thanks to her!)

"Proper" statement is "Say something this way when you are talking to those representatives of the state that insist on obsequeous behavior towards authority". Yeah. And I do. But I know better than to take their demands as any description of truth. Or justice.

I am ..... the LAW!

[mustangdavis]

This kind of technology sounds kinda scarey ... just like in the (horrible) movie, Judge Dread. Yes, the technology could help police and law enforcement to solve crimes that would otherwise possibl be unsolvible, BUT is it worth it? I see the opportunity for a very hatefull person to set up someone for a crime they didn't commit ... and for the evidence to be used against them in court! Yes, this does exist today, but atleast there is a CHANCE that the person falsifying the evidence COULD get caught. How could this person get caught if he is permitted to MAKE evidence?

Its a good idea, but this DEFINATELY leaves a resonable doubt as to whether the evidence is real and legit ... which could also be used AGAINST the prosecution in what would have otherwise been a solid case ....

Bottom line: too much risk for too little benifit.

Re:I am ..... the LAW!

[Dusabre]

How is this capitalised mess 'insightful'? The police aren't asking for the power to 'make' evidence (They can make 'evidence' anyway if they want to break the law - traditional images and physical evidence can be and is tampered with). They want to use advanced filtering and enhancement techniques on digital images.

In principle these are not different to blowups, negative colour and other standard analogue photographic techniques that are already used in court in relation to photographic evidence.

If an audit trail is kept, I don't see a problem with using these techniques. Digital manipulation is a powerful tool that can be abused but as long as the source is kept pure, the manipulation can be exposed.

There is another problem with fingerprints though, the science behind fingerprinting has never really been proved. Do a google if you want to learn more.

Re:Use a real verb

[L. VeGas]

Just google a bit and you can see he's right.

get it? google?

haa haa ha ha ah never mind

Useful... possibly

[rworne]

The only way this can be seen as useful is if the person who is "cleaning up" the fingerprint has no idea who the print belongs to and where the print came from. Considering all the prints the law enforcement must deal with, it would be hard to assume the print a tech is working on is for a high profile murder suspect or a car thief.

That way it removes the ability to "doctor" prints to match what the cops want, and it adds a valuable tool to the investigative process.

If this process involves the tech working on a print, with the "target suspect" print available to him, I'd cry foul in an instant.

Re:Useful... possibly

[EvilBudMan]

--The only way this can be seen as useful is if the person who is "cleaning up" the fingerprint has no idea who the print belongs to and where the print came from.--

Someone MOD that man up. That would be the only safeguard that would work.

Re:Useful... possibly

[cosyne]

The only way this can be seen as useful is if the person who is "cleaning up" the fingerprint has no idea who the print belongs to and where the print came from.

Or, you could RTFA.
"I don't think I could recreate a fingerprint," said Knoerlein, pointing out that he never sees the suspect's fingerprints. The system might be more vulnerable where print examiners have both sets of prints and also are responsible for enhancing the prints, he said.

Re:Useful... possibly

[HiThere]

A real problem is "How do you tell one digital output from another?" I can imagine ways, and for most ways I can imagine, I can eventually imagine a way to defeat it. And are any of these approaches being proposed?

That the official tech didn't know what he was changing to or from is only one level of protection. Then you need to ensure that his copy of the file is the one that ends up in court, but computer files can look quite remarkably alike, unless quite careful precautions are taken.

Not Digital Rights Management!

[Cranky_92109]

This not DRM at all, what they want is a system whereby they can say that a digital photograph has not been altered. So yes, this is very important. Imagine you're on trial for a crime you did not commit yet the prosecuting team has altered a similar fingerprint (but not similar enough for conclusive evidence) to make it more like yours.

DRM is an important fight, but keep the boundaries clear or you'll muddle your argument. These are separate technologies.

Re:Not Digital Rights Management!

[SuiteSisterMary]

Well, that's easy enough; make 'digital film.'

By that, I mean write-once PROMs that can be popped into a digital camera; they can hold x number of pictures, and when you click the button, the picture gets burned right into the PROM.

Burn in a checksum or something as well, and you can tell if bits were removed. Build it so that it's not random access, and you can't swap bits around or anything.

Sure, you'd need another system if you WERE doing enhancements or changes or anything, but the ability to pull out a PROM chip and say 'here is the original photo, guarenteed unaltered' would be good.

Re:Not Digital Rights Management!

[kreyg]

By that, I mean write-once PROMs that can be popped into a digital camera; they can hold x number of pictures, and when you click the button, the picture gets burned right into the PROM.

Except... if you take a picture with a digital camera, Photoshop it, and then burn it into such a PROM from the computer (which should be trivial), you've only gained the illusion of authenticity.

Actually, I kinda like this idea

[unfortunateson]

...and I'm far from liking DRM in most cases. But with digital images so easy to manipulate, I need a trackable, reliable image for: a) Insurance claims -- So the insurance company won't declare that I hacked the 60" plasma TV onto that blank wall, then claimed it was stolen b) Reliably submitting news/crimestopper photos, so that they can't be debunked c) Plagiarized photos can be tracked to their sources This isn't a panacea, and I would certainly want to be able to not mark some images. The audio environment currently has issues where you can't dup your own personal recordings -- this is wrong. But I should be able to fingerprint my own media, and declare its copyright.

No.

[dacarr]

Fingerprints are unique per person. Any subtle change could incriminate the wrong person.

Re:I fully support DRM cameras

[tempest303]

until then, DRM may be that only thing saving our marriages.

*giggle* My, you're a such a silly little troll! Teehee!

Re:I fully support DRM cameras

[Kallahar]

Uh, DRM wouldn't help that, I can give you the culprit right now: THE HUSBAND.

How the hell did you get modded up?

Travis

Re:I fully support DRM cameras

[L. VeGas]

Can you imagine the shame of having a co-worker, or postal clerk, or supermarket butcher comment on how good you look in a negligee?

Gee, it would be kind of flattering actually. I think lavender would nicely offset my big, black beard.

Tween the image and voila, it's anyones fingerprin

[BoomerSooner]

Stupid really. Has anyone seen those videos where they tween one face to another?

What's to stop law enforcement from doing this to a fingerprint?

Remember we have the LAPD (planting guns/drugs), NYPD (broomhandle, Central Park Rapist(s)), ...

When a DA is freaking out and the public is wanting someone caught, the stresses of these situations generally lead to bad things for innocent people. Plus if you're a minority it's significantly more likely that you'll number one, be convicted and number 2, face a significantly harsher sentence than a white/anglo counterpart.

How to stop this method from working in court...

[Brightest Light]

One of the biggest questions about the new technology is: Could a skillful technician create or copy a suspect's fingerprint and frame someone by making it look like that fingerprint was at a crime scene?

Here's an idea: Get a copy of the print image, find somebody with Photoshop skill, get them to alter the image to show judge's/prosecuting attorney's/etc's print; evidince (hopefully) supressed when the judge realizes how easy it is to fake 'evidince' that way.

Re:Use a real verb

[rickliner]

This concern has been expressed more eloquently by a young but wise voice:
"Verbing weirds language." -- Calvin, to Hobbes

Re:I fully support DRM cameras

[BabyDave]

I was with you until I read this bit:

Ideally of course, the digital camera would have an AI that would prevent taking pictures of naked women, but until then, DRM may be that only thing saving our marriages.

That's a different argument entirely - it's saying that the "wrong" act was taking the pictures, when previously the "wrong" act was distributing them. If people want to take naked pictures of each other, that's their business and no-one else's.

If you'd said "... would prevent taking pictures of naked women unless they truly consented ..." or something, I might not have disagreed with you (possibly I would have pointed out how difficult that would be to judge).

Not the cameras, the subjects !

[stud9920]

The DRM should be not be on the cameras, but on the subject ! According to the LAMAA (Landscapes And Monuments Association of America), five billion pictures of copyrighted monuments and landscapes are illicitly taken each year. (actually it's three billion pictures, but some of them are printed on posters, and Monument Valley counts as two, being a monument AND a landscape).

No royalties are ever paid for these pictures. Some hippies are claiming "fair use" because they paid the entrance fees to the park/monument.

A big part of them are shot with a "friend" in front of the monument/landscape. By the use of such circumvention devices, the photonic pirates claim they are creating a new work, supposedly protected itself, in fact pure piracy. Such a circumvention device should be outlawed.

Therefore, the LAMAA (Landscapes And Monuments Association of America) demands that the SSSCA be amended to make DRM on new monuments and landscape mandatories. Such a device would render all but these so-called "friends" black on the photographic device and thus encourage the fine LAMAA members into providing exciting new monuments and landscapes, like the upcoming Senator Hollings Memorial, the Shores of Montana (thanks to our president's Kyoto enforcement), or the new World Trade Center.

Any evidence can be tampered with

[stratjakt]

Thats why the term is 'reasonable doubt'. The DNA technician can lie. The blood can be planted. You can doctor an analog photo just as easily.

I've mentioned before that I design, write, and support police records software. I know how important audit trailing is to them, I was up until 3 AM last night debugging some of it.

We've even been approached with this very idea, audit trailing and securly storing digital photos. (Not just fingerprints)

This is about showing a factual list of who had access to the photo, exactly what they changed, and when. If pixels were added, it'd be on the trail. If it was lightened, darkened, it'd be on the trail.

The reason is simple. The jackass lawyers who think the constitution spells out their job as 'get the client off, no matter what it takes'. Another rant entirely, but rigorous defence doesnt mean knowingly lying and misleading a jury.

Police are constantly accused of lying, tampering with evidence, planting evidence, in stupid cases like misdemeanor posession of pot.

So when Mr Defense gets up in front of the jury, with Mr Cop on the stand, and says "Isn't it true that anyone could have altered those photos?", "Mr Cop can say, here's an itemized list of every enhancement, change, and view of the photos since they were taken.".

If Mr. Defense is stupid enough to continue, they can present sworn depositions from people like me (who created the system) testifying to the authenticity of the data.

Of course - it'd go both ways. If Mr Defense truly thinks BeatCop O'Malley doctored the photos, someone like me could likely prove the when and how.

This isn't a bad thing, or about stripping rights. It's about helping to secure the right to a fair trial.

Re:Any evidence can be tampered with

[el_chicano]

I've mentioned before that I design, write, and support police records software. I know how important audit trailing is to them, I was up until 3 AM last night debugging some of it.

So you would be the first to admit that your opinions are colored due to the way you earn your daily bread?

The reason is simple. The jackass lawyers who think the constitution spells out their job as 'get the client off, no matter what it takes'. Another rant entirely, but rigorous defence doesnt mean knowingly lying and misleading a jury.

Knowingly stating an lie in court is perjury. I am sure that if your DAs are compentent they could convince a judge that it is perjury.

Also, show me in the Constitution where it says that defendants are not entitled to the best defense they can afford. Or where it says that defendants are guilty until proven innocent. Or where it says it is OK for the cops and DAs to lie.

Police are constantly accused of lying, tampering with evidence, planting evidence, in stupid cases like misdemeanor posession of pot.

And cops never lie? Shit, in Houston, Texas the HPD busted down the door of a Mexican citizen and assassinated him in bed. Oh, he was a "drug dealer" but no drugs were ever found. That was a FATAL lie by the police.

If pot cases are "stupid" (and I agree with you there) then the stupid law needs to be changed. But just try to legalize pot and see all the right-wing pro-law-enforcement groups come out with the Big Lie against weed, led by the Doofus in Chief in the White House.

Don't you know that marijuana kills? Or when you buy weed you are supporting terrorism? The really sad part is the sheep, err citizens, willingly believe this crap!

Every photo is "enhanced"

[acomj]

Even when you print negatives with a standard darkroom enlarger, you adjust contrast/ exposure / color. You can highlight details, darken/lighten areas etc. You can do alot with standard darkroom equipment.. (waterfalls down escalators etc.)
Even printing slide film involves adjustments.

With the computer and photoshop its even easier to "adjust" photos.

Digital watermarking of images already exists. I don't know how effective it. If images by cameras are watermarked this may image authenticity. But will watermarks those survive jpeg converstion, they can be faked too. Many jpegs from digital cameras already have information about camera/time/exposure imbeded in them.

I don't think there is any way to trust photographs. Look at the fakes with analog (ufo shots / Loc ness monster etc..) With digital it just gets harder to believe.

DRM won't help in that case.

[Frobnicator]

After READING THE ARTICLE the suggestion that using DRM or encryption doesn't make sense. They are modifying the image by running various filters and kernels over it, trying to enhance and draw out information. Additional protections of the file does nothing to protect the image integrety. What they are having a problem with is using the enhanced versions of the prints.

The only thing the DRM or encryption would do is provide yet another means of tracking the files -- but it sounds like they are already using safeguards there. All versions, the user, and the duration of use are tracked. Those are the same, or in some cases better, than protections of physical evidence.

They don't need DRM cameras or higher cost encryption schemes. They need the same arguments that first allowed for fingerprints, DNA testing, and other new technologies in the courts.

frob.

thoughts..

[Dr. Awktagon]

some thoughts..

* Anytime you use encryption or digital signatures, it's not "DRM". It's not like these folks want to restrict copying of the pictures, or track people who see the photos, they do that by keeping the pictures within their labs. The encryption is so they can show in court that the picture was not tampered with. When you check the signature in your linux package files, that's not DRM, that's something for your own benefit.

* I was recently looking at Canon's latest EOS-1Ds camera, which has a "Data Verification Kit" encryption module available. You put a smart card in a reader and every shot is digitally signed in the camera. So this stuff is available and hopefully the forensic photographers will begin using it. Of course a malicious photographer might change the software in the camera somehow but hopefully the module checks for that.

* Dodge & Burn tools should probably *not* be used.. they allow you to darken/lighten specific *areas* of the photo, which could be dangerous. When enhancing evidence they should only allow *global* changes like overall brightness or contrast, etc. Or at least they should send the evidence to three or more independent labs, who don't know anything about the case, and let each version be seen in court. That way there's less of a chance that someone will doctor the evidence for a specific outcome. And of course the whole workflow needs encryption and signatures.

* Evidence can always be tampered with. The digital signatures just make it harder, and hopefully at least as hard as it is now in the non-digital world.

This is not how photographic evidence works

[Bruce Perens]

The potential to fake a photograph has existed since the earliest days of photography. The veracity of the photograph or other scientific evidence rests on the oath taken by the photographer and other technicians involved. They are not asking for DRM, but for a digital signature generated by the camera and attesting to the time and other circumstances of the photograph. Auditing the image-processing is possible, given certified software and a circumscribed list of permitted operations. Some form of "trusted computing" could be used to avoid trivial circumventions of the list of allowed operations.

Note that this is "trusted computing" in service of the owner of the computer (in this case the police department and department of justice rather than the individual operator). The fundamental difference is that the owner of the computer is the one asking for the trusted service, rather than some other entity that does not trust the owner of the computer.

Bruce

Re:This is not how photographic evidence works

[BWJones]

The potential to fake a photograph has existed since the earliest days of photography.

This is certainly true and for a decent review, check out Dino A Brugioni's book, Photo Fakery: The History and Techniques of Photographic Deception and Manipulation. ISBN: 1574881663

Dino served as one of the CIA's senior photographic interpreters and the book is a decent review.

As far as DRM and how digital image management could work, see my other comment regarding the solution to DRM issues and certification of images.

Professional point of view.

[JPhule]

Speaking as a member of the Law enforcement community, I see how increasingly difficult it is to get a good solid conviction.

This is just playing with fire. The encrypted cameras sound like a good idea, but I think that you need to have more solid evidence. Video cameras in squad cars is a great example. When you can get a drunk to admit how many beers he has had on video tape while conducting a field sobriety test it is pretty easy to refute his claims in court that he was just driving home from grandma's house and got a little tired.

The thing is, maintaining a trail of custody for the photos I think would be much harder, therefore easier to refute their validity in court. And any time you start messing with anything remotely related to being circumstantial evidence, you might as well just throw the case out the window, cause thats just what the judge is going to do.

It is going to be hard to convince people that this is a technology with feasible use in the courtrooom after they have seen pictures of OJ wearing ducky slippers.

Are you serious?

[kabir]

At first I thought this was a humor post ("AI that would prevent taking pictures of naked women"?!? I mean really.) but then it occoured to me that you might actually be serious. So, for just a moment, I'll step back from inflamatory language ("sick-men", "villany") and silly ideas (the aforementioned AI) and try to address your comment.

What you're proposing is that there should be a technological to what is, at it's heart, a moral problem. If you're looking at this from a moral/religous standpoint (and it sounds to me like you are) then you must recognize that you haven't solved anyting by taking away the option to engage in this sort of immoral behaviour (I'm assuming, for the sake of the discussion, that audit trails would, in fact, stop this sort of breach of trust from occouring - though in all honesty I think that's a deeply flawed assumption). Certainly the specific behaviour might have been prevented, but the underlying issues which allow someone to ignore or abuse their spouses trust, etc. still exist and have, in fact, been completely ignored. In the end all you'll end up doing is changing the specifics of the situation, but breaches of trust and sexual compulsion will continue. No amount of DRM can address the reasons that someone seeks out bad behaviour.

Of course, as I mentioned, DRM and/or audit trails won't do much to stop the unauthorized publication of such photos. It certaily won't stop one spouse from coercing/manipulating/etc. another to take the photos in the first place. In no way will it account for those who initially think the photos are a good idea but later change their minds. And finally, it's complete bunk to even begin to think that an audit trail will force anyone to "confront thheir odious addiction", and it certainly won't save any marriages.

The bottom line is that if a husband thinks so little of his wife and their relationship that he would violate her trust in sunch a way simply because he wouldn't get caught then the marriage is in serious trouble regardless. Technology is amazing stuff, but it's neither the cause of, nor the solution to each and every modern problem. I think people forget that too often.

Zero Post ??

[HughsOnFirst]

That's funny, I posted a link to the appropriate info yesterday, here it is again

Here is a good article that covers a lot of this
The "Authenticity Crisis" In Real Evidence
Scientific Evidence Review
10.1.2001

You might also be interested in the KODAK Picture Authentication Module [kodak.com] which uses PKI in a camera.

If I post before the story goes up is that a "Zero Post "?

Having been involved in traditional analogue photography for 30 years, I can tell you that I'd trust one of those Kodak cameras more that say a 35 mm Ektachrome Transparency, or worse yet, a color print. A while back Polaroid was blowing out a digital printer that output on spectra film for 30 bucks. I considered buying it for all sort of practical jokes and parking ticket disputes.

Seriously...

[Eric Damron]

Okay seriously. The problem with digitally enhancing something is that depending on how you "enhance" it you can make it out to be almost anything.

For example: The "face on mars." Enhanced one way and it looks like a face, enhanced another and it's just an unimpressive hill.

Digital certification

[BWJones]

The real solution here is to create a national database for digital imagery of evidence through the FBI most likely. When you capture a digital image, that image is then sent to the central repository where the image is "fingerprinted" and registered. This way images can be certified as either original, enhanced, or tampered with when the issues come up in court.

Photoshop Vs. Law

[Malicious]

I Photoshop my Photo-radar tickets, superimposing the police officer's liscence plate in place of my own. I then take the edited ticket, to the Courts, and claim that this is not my liscence plate, and the charge is thrown out.
Photographic evidence, specifically such evidence, that has clearly had the oppertunity to be edited, should never be permissable.

Evidential material, or not?

[bagofbeans]

I use to deal with a company in UK who handled remore control and transmission of real time video from security cameras and similar.

They told me that digital material is only treated as 'evidential' by UK courts if it has not been processed digitally any way. Raw data from digital cameras is ok, but lossless compression (zip, rar etc) cannot be used before storage or transmission of data. Thus encryption is not allowed, and lossy processing like MPEG? Forget it.

This is one reason why security recording are still largely analog - VCRs. Another reason is that VCR tapes are cheap, hold a hell of a lot of information, don't take up much room, and can be re-used.

It is probably fairly easy to present reasonable doubt of digitized evidence unless the resolution is so that tampering would be detectable.

encryption != DRM

[Twirlip of the Mists]

Guys, encryption and DRM are not the same thing at all.

What's needed here is a "tamper-proof" digital image format, one that can't be modified or that can't be modified without leaving a record. Think checksums and digital signatures here, comprehensively applied. The same thing will be useful not only in criminology but also in medical imaging and lots of other areas as well.

DRM has nothing to do with "tamper-proof" data. DRM, which stands after all for "digital rights management," is simply a catch-all term for any technology that serves to capture rights as metadata, and possibly control access to media according to that metadata.

As I've written before, DRM is most important in the commercial TV broadcast space. A TV station buys a "rights package" for a syndicated program, and has to pay a very large fine if they violate the terms of that package. (Say, if they show the program at 10:00 AM when the contract says they can only show it between noon and midnight.) DRM in that arena will be a life-saver for those kinds of folks.

I know this is Slashdot and ungroupthink is doubleplusungood, but DRM is not a dirty word, and DRM and "tamper-proof" media are not the same thing at all.

Yes to both questions

[bmetzler]

Law enforcement agencies need secure cameras because criminals are slimy and judges often side with criminals. If a criminal can claim that evidence should be thrown out because it wasn't secure, you bet the judge will side with him.

We also need the right to "photoshop." Enhancing a photograph is no different then using a microscope to make the image more clear.

Although I could compromise on the first one if the law would be strong enough to prevent evidence from being thrown out, we need a strong law to be passed allowing digital enhancing.

-Brent

Re:Yes to both questions

[dvdeug]

If a criminal can claim that evidence should be thrown out because it wasn't secure, you bet the judge will side with him.

Right. The public defender who writes The Law is a Ass says that he's near the top of his office having got evidence thrown out twice, and that judges hate to throw out evidence just like anyone else would.

Enhancing a photograph is no different then using a microscope to make the image more clear.

Once you put something under the microscope, you can play with a couple knobs, but what you see is what's there. There's no such guarentee with Photoshop -- you could load it into Photoshop, erase it, and draw a new fingerprint. "Enhanced" photographs shouldn't be accepted until it's passed the scientific acceptance test (like microscopes have), and even then it's important to tell what was done to a picture.

Re:Yes to both questions

[el_chicano]

...If a criminal can claim that evidence should be thrown out because it wasn't secure, you bet the judge will side with him.

...Although I could compromise on the first one if the law would be strong enough to prevent evidence from being thrown out...

Doesn't anyone read the US Constitution anymore? It is intended to LIMIT the govenment's power and ENHANCE the people's power.

I do not see how in the world this post was modded as insightful. It is just parroting the same old tired right-wing "law and order" rhetoric. Probably the mod was some "Ditto-head" or he/she works as a cop or DA...

Yeah, lets give our cops and DAs <insert random power here> as long as it is rigidly controlled and fair and just. Trust us, the government doesn't make mistakes.

Look at the death penalty. The overwhelming thing that jumps out at you is that minorities are on death row disporportionate to their population percentages. There are also more poor people on death row than rich people.

For some reason, White techies tend to be conservative, they are always dogging the liberals and espousing Republican and/or Libertarian viewpoints. Why is that?

Look around you, it is the Republicans that are taking rights away from you. They are selling this nation down the river but as long as you White techies are making money you are content to ignore what is going on around you.

Bah! I am getting disgusted with this country and am looking for somewhere to go. I'm thinking Costa Rica, Spain or maybe even Ireland. Anyone have any suggestions on any techie nirvanas out there?

Re:Yes to both questions

[bmetzler]

There's no such guarentee with Photoshop -- you could load it into Photoshop, erase it, and draw a new fingerprint.

You must think that the justice system is full of idiots. I'm sure that this "new" fingerprint is going to hold up against the old fingerprint really well.

-Brent

Re:Yes to both questions

[dvdeug]

You must think that the justice system is full of idiots. I'm sure that this "new" fingerprint is going to hold up against the old fingerprint really well.

Erasing it was the extreme example, but one of the big problems with our justice system is "experts" spewing nonsense and juries buying it. If a fingerprint expert gives us an extremely-blurry before and a clear after, most juries are going to accept the change on his word, unless it's totally wrong. The jury is not likely to be packed with fingerprint experts; they won't know what was significant in the original and what was just noise. Is that swirl part of the original design, or did the object turn while the perp was holding it?

Re:Yes to both questions

[bmetzler]

The jury is not likely to be packed with fingerprint experts; they won't know what was significant in the original and what was just noise.

Defense lawyers generally don't roll over when confronted with things like that. I'd expect that before the trial was over, the jury would hear all about the possible problems.

-Brent

Re:Yes to both questions

[dvdeug]

Defense lawyers generally don't roll over when confronted with things like that. I'd expect that before the trial was over, the jury would hear all about the possible problems.

Well, gee, then we should just let anything in. Not everybody has O.J.'s defense lawyers, and even with the best defense lawyers, an "expert" can be a pretty condemning thing, no matter what a defense lawyer (sleezy scumbag defending murderers) says.

Crypto-fingerprinted JPGs

[jridley]

Kodak has had a mod available for their cameras for a long time; at least the DC290 used to have it, several years ago. The firmware embedded a crypto fingerprint in the EXIF header, and they had a piece of software you could run to read it back and verify that the JPG, exactly as written, was created on X date in Y camera with serial number Z. It was intended for just this use.

DRM?

[Audacious]

It doesn't (as far as I could see) mention DRM. It only talks about being able to follow who did what with the information.

However, even the system of encryption, et al which is being proposed doesn't really do much. First, is the machine picking the randomly generated password or the person picking an easily recognizable password as in this problem.

Second, the machine (whether it be a camera, computer, or nightstick) doesn't have intelligence built into it. Thus, it would allow anyone who knew how to work the machine (and could guess the passwords) to alter the information. Even the fact that the computer is smart enough to make a copy of the original doesn't mean anything. If someone knew how the program worked - they could (and would) alter the original as well as the copy.

Until machines become self-aware or at least are aware of what someone is trying to do to them - we will not have a "good" way to stop fraud. (I say "good" because even then we will probably have some way to circumvent/unplug the smarts from a machine which puts us back where we are currently.)

Thoughts:

If the police want a more fullproof method of maintaining equilibrium in the establishment of, and verification of proof. Then they will need to greatly improve how that information is handled. A network (probably made up of Linux boxes) which are attached to a central repository and to which they can send information but not retrieve information (ie: a blind send) would be a step in the right direction. Information would only be retrievable from the main console connected directly to the centralized hardware. Also, files can not be deleted from the main system until the files have been backed up to a reliable medium (such as CDs/DVDs/tape). Otherwise, the system simply allows a user to register updates and nothing else.

Contrast

[limekiller4]

From the article:
"'I think it's very suspicious that you have something that is of no value and suddenly you enhance it and it becomes of value,' said [defense attorney] Heyer. 'It is very clear that this type of thing can be manipulated.'"

I dislike it greatly when a person states a truth then follows it up with an obvious untruth in order to give the latter credibility. Yes, evidence can be manipulated. yes, it should be protected against. No, being able to discern something you could not discern before does not invalidate that evidence out of hand. Perhaps she's heard of DNA?

I think just about anyone familiar with Photoshop/gimp and actual photographs will realize that details can be brought out of a picture that aren't immediately obvious. Don't believe me? Take just about any non-perfect picture off the net, open it with gimp, then:

• right-click on the image.
• Image > Colors > Levels
• Tweak settings.
• Profit! (just kidding...)

Should it be held under the cold light of courtoom examination? Sure. Is it pseudoscience? Not on it's face.

Re:How to stop this method from working in court..

[LittleGuy]

Let's not get our collective panties into a bunch, people. 'Framing people' existed long before Vic Mackey picked up the local Yellow Pages to interrogate a suspect and Jack Bauer found another use for a bowling bag.

This is case of a another facet of technology that can be used by a Corrupt Offical For Nefarious Gains[TM]. If it exists, it will be used. And it will be allowed only when those in judgement allow it to be used.

Threshold

[limekiller4]

I think this is a no-brainer. If law-enforcement wants this to be taken seriously as a tool (and I'm all for the theoretical good guys having better tools to make sure the bad guys stay out of the general populace despite what they'll tell you about us criminal-loving liberals), then all they need to do is demonstrate through double-blind trials that use of the tool does not lead to an increase in false-positives.

Simple.

As for it's potential for abuse, give me a break. Planting a print at the scene is about a kabillion times easier to do than to digitally forge one. Occams Razor, kids.

Re:Use a real verb

[volpe]

Photoshop is a trademarked proper noun, not a verb

Any noun can be verbed.

Keep Film Cameras

[mbstone]

In the People v. O.J. Simpson case, Simpson's lawyers demanded the contact sheet from the police crime scene photos. They didn't get the contact sheet -- which conclusively reveals the order in which the photos where taken -- until the trial was nearly over:

THE COURT: The court will entertain a motion to shorten time. All right. Any other Brady issues we need to discuss?

MR. SCHECK: There is the matter of the contact sheets.

THE COURT: I'm sorry?

MR. SCHECK: The contact sheets.

THE COURT: All right.

MR. SCHECK: We believe, your Honor, that the testimony yesterday of the other--this week of Mr. Rokahr is extremely significant. It is significant because it establishes that the picture of Detective Fuhrman pointing at the glove occurred at night, not at 7:00 in the morning or about an hour and a half after daybreak. More significantly, it puts Detective Fuhrman in an area by the evidence at a time when he is unsupervised or unobserved, I should say, by others, and it is in complete contradiction not only with his testimony, but an apparent contradiction with the testimony of other officers. The pictures speak for themselves and are extremely powerful evidence.

Now, the problem that we've had in this case, as the court is well aware, is that we have been requesting, A, a photo log or a listing of the order and/or time that pictures were taken, or B, an opportunity--or contact sheets, or C, an opportunity to make contact sheets to the negatives. This was pursued in discovery in pretrial and it was pursued during the trial. Now, Miss Clark got up here yesterday and said that she was able to look at the photographs and see numbers on the face of the photograph and she herself had put together a stack of the photographs in order so that she knew the sequence. Well, that is something that Dr. Lee, Dr. Wolf, Dr. Baden, myself, the other lawyers on this team, could not do for a number of reasons. No. 1, there are different photographs--photographers in this case and the numbers--there were more than one, for example, picture no. 35. No. 2, as the court noted and Mr. Rokahr noted in some of these pictures, the--you can't see the number. And most important of all, your Honor, as the court noted itself yesterday, the best proof, the real proof as to what pictures were taken in what order are the numbers on the Kodak print. That is the real proof. That is the hard-core proof.

THE COURT: Frame number on the film.

MR. SCHECK: The frame number on the negatives. Now, the problem I have, and I mentioned it to Mr. Hodgman yesterday, is that Mr. Hodgman is the lawyer and then later Mr. Yochelson, who was designated to be dealing with discovery for the Defense. And this court is aware, and Mr. Hodgman has conceded candidly with this court, that they had no photo logs, they had no ordering of pictures, that they did not believe, until we discovered this contact sheet from Bodziak, that these contact sheets could be made from the negatives, that there was no ordering. I am very, very troubled by the fact that Miss Clark is telling this court yesterday that she had a list of photographs and an ordering. Well, if she had it and they were determined--they had determined--she had determined through conversations with photographers or detectives on this case, that she knew which pictures were taken in which order, then we are being misled because we are only dealing with Mr. Hodgman and Mr. Yochelson who know of no information in the District Attorney's office or in law enforcement that contain an ordering of the photographs.

Unenhanced fingerprints also unreliable

[spun]

This articleoriginally from the New York Times, has some interesting information about fingerprints. Even without photographic or electronic enhancement, fingerprint identification is being called into question. Most experts in the field are self-proclaimed and self trained. The field of fingerprint identification is completely unscientific.

Limited photoshop?

[gclef]

Why not just create a version of Photoshop that can only do things like contrast, burn, etc? Remove the tools that allow image modification entirely, and sell it to police forces as a way to get around this problem.

I see a software niche....

Enhancing Evidence

[cosyne]

Image transforms do not add information to an image, they just make it easier to see the information which is there (try using Photoshop Auto-level to make an image of bill clinton shaking hands with an alien). Using dodge and burn over an entire image or a large area of it will not let you change fingerprints, just make existing ones easier to see. However, if you go into photoshop and use a one pixel burn brush you could draw lines with it. This is why it's important that the person doing the image processing isn't also doing the fingerprint analysis. It's like medical imaging- the imaging tech generates a good image, the doctor decides what it means.

As for the 'As if by magic' and 'psuedo-science' bits in the article, those are irresponsible hype. It's like saying you expose film in a camera, develop it, and an image appears as if by magic. If you didn't know how a TV worked, you'd think that was magic too. As for the unrepeatability of results, no two people using fingerprint dust will get exactly the same results. Same with a photoshop brush. If you brushed the same areas in the same ways, you'd get the same results, otherwise not. Duh.
This does bring up a point of repeatable, localized image processing. My guess is it wouldn't be too hard to get the GIMP to record all brush strokes. It surely stores their results for the undo option. How hard would it be to output an XML encoded series of operations along with the output image? Then if there's any question as to the usability of the results, someone can start with the original image and apply the same set of operations one at a time. Maybe add image cryptosigning, and sell linux+gimp boxen as forensics tools.
Finally, i'm surprised there isn't a standard government issue image transform system. NIH Image might be a good place to start, or just a front end to matlab's image processing toolbox which is luser-friendly and keeps usable, crypto-signable records of each transform it does. As long as there aren't any brushes, no expert witness in image processing is going to say you could doctor prints.

Re:Is Photoshopping a word???

[Twirlip of the Mists]

I'm pretty sure what makes it a word is having the other person or people understand it without explanation. Even if you don't know what Photoshop is or what one does with it, the meaning of "photoshopping" can certainly be glorked from context.

Get the right tool for the job

[Melantha_Bacchae]

First off, IANAL, and IANAForensic expert of any sort. I have used photoshop for years. And I've worked on software projects requiring government certification. These are my opinions.

That said, DRM does not apply (you are certifying the source here, not caring about end user rights). Photoshop is way too broad a tool, with too many abilities to create your own content. As for a digital file, don't put that in your mouth unless you know where all it has been.

What you would need would be image processing (not editing) tool, preferably specific to enhancing fingerprints. The best thing would be a self contained fingerprint enhancing appliance, with scanner, printer, and built in algorithms. The fingerprint would be scanned in, enhanced, and then go back to the real world as a watermarked print that could be taken to court with the device's serial number and the original fingerprint.

The device would of course be fully certified to do exactly what the court would admit. And that is the ticket: you need a fully controlled process that can be examined at every step with a fine tooth comb by some agency of the court to prevent forgery of evidence. You also need to link the evidence to the specific machine, so it can be hauled into court and publicly verified that it hasn't been tampered with.

Of course, to make Slashdotters happy, the device could run embedded Linux, and use Gimp routines, as long as you could find somebody to fill out enought paperwork to keep a certifying agency happy. A few boxcars would do. ;)

This device would not be a consumer or pro graphics device. So there is no need (or even desire) to burden the public and the pro graphics community with the requirements of forensic evidence.

"The thousand year dragon king: King Ghidora."
Yuri, "Godzilla, Mothra, King Ghidora: Giant Monsters All-Out Attack"

Don't forget King Ghidora's birthday is tomorrow.

Re:Tween the image and voila, it's anyones fingerp

[Qrlx]

Of course, you don't need Photoshop to fabricate evidence. It's just another tool that can be used.

A lot of people have complained that who knows what the authorities might do with Photoshop -- enhancing evidence and such. That's a valid point but you should take a step back and realize how scientifically flawed fingerprinting is in the first place. (In my opinion, of course.)

Fingerprinting came about around the turn of the twentieth century as a replacement for a failed biometrics system, in which certain mesaurements of a person (size of head, length of arm, stuff like that) were being tabulated, and recorded to make a database of known criminals. Problem is, two people could have the same measurements.

Likewise, there is no "guarantee" that two individuals have the same fingerprints. Observation has shown that two people probably don't have the same prints, but that's no guarantee. I don't believe the medical community even really understands what makes fingerprints "grow" in the first place.

Fingerprinting is not a "science" in the way physics, chemistry, etc. are. (Legally, this is called the Daubert Test.) Where is the peer review? If fingerprinting were truly a science, as American courts have determined science to be, the national fingerprint database should be publicly accessible. It is not. The formula/algorithm by which fingerprint examinters determine a "match" would be public. The method that the computer uses to match fingerprints would be public knowledge, but it is not.

I'm not trying to say that fingerprinting doesn't provide valuable evidence, and I certainly do believe that fingerprint evidence is a good indicator that somebody touched something. But is it iron-clad proof? No. And worse than that, is is a closed-source, proprietary system.

Were fingerprinting evidence to be invented today, the courts would probably not allow it. It has not withstood (likely it cannot withstand) the same sort of scientific scrutiny that DNA identification has. However, they have significant enough momentum behind them that even though they may not be an "exact science" they are good enough for the purposes of the criminal justice system.

Here are some good links:
Federal Judge Slams Fingerprint "Science"
Cornell News: Fingerprint Study
Latent Print Examination disagrees with most of what I say...Click the Ressam link...if you don't support fingerprint evidence, then you support terrorism!

My favorite quote

[SeanAhern]

My favorite quote from the article:

The software used to enhance the print is the same that some tabloid newspapers use to create seamless "photographs" of space aliens hanging out with celebrities. Time magazine used a similar program to alter a police mug shot of O.J. Simpson and make his complexion appear darker on its front page in 1994.

Heyer said that shows the software is not a scientific tool, but an unreliable art form that could be used to misrepresent reality or simply create things.

Lemme see if I got this straight... Since photoshop can be used to fake photographs, that means that it can never be used as a scientific tool.

Should we make the same claim for film development equipment? What about scissors and tape?

I'm all for more accountability when it comes to evidence. Having an audit trail for all digital processing is a wonderful solution. That way you can always verify who did what to an image, from start to finish. Heck even if an analyst uses a tool like dodge and burn that requires "painting" onto an image, you can keep a record of the trajectory of the pen stroke used. This isn't rocket science here.

But to claim that a valuable tool of science, because it can also be used to create art, is suddenly invalid, is simply reckless and irresponsible.

Since when did a product name became a verb ?

[forged]

So I guess I must BMW to work every morning, and ask my customers not to Nokia me in the middle of the night. Clever !

Perhaps the verb you were looking for was to enhance ?

Nonreputiation, accountability

[theonetruekeebler]

Anybody remember the OJ murder trial? One of the major tactics of the defense was to make it sound like every cop in L.A. county had handled---and fucked with---the evidence. In the hands of a lawyer, Photoshopping a picture to sharpen the contrast is "altering the evidence using sophisticated image-manipulation software". That doesn't sit well with a jury, good intentions or not.

In a trial, the chain of possession on evidence must be sound. You must be able to demonstrate who has had the evidence, and what they might have done to it.

Think of this as a tamperproof RCS for photographs: You have proof of who has touched the evidence, and what they've done to it.

Data Mining, "Airframe" and more.

[aWalrus]

In the book "Airframe" by Michael Crichton there is one part where they have really blurred footage of an accident aboard a plane and the main character requests an enhanced version of it. The technician suggests some techniques that would be based on taking educated guesses, but she refuses and tells him to only perform totally computer based enhancements to it. They end up running the frames at high speed and taking "snapshots" of the composite result, which greatly enhances the details.

So, what's that got to do with anything? The use of enhancing techniques CAN be abused. By selectively enhancing a section of a print against another one you can manipulate the outcome. It is very different to increase visibility on the whole print than to darken some parts, blur others. If there is a human element, it is not to be trusted. Allow me to explain (and sorry for the long rant):

Some time ago I attended a Data Mining Seminar. Many people here will be familiar with the techniques that are used in that area. What's the point of using a computer algorithm to find patterns when we are so good at it naturally, you ask? Why, exactly that. Humans are extremely good at finding patterns. We even tend to see them where they're not, and ignore the extra evidence that may point in different directions. A computer works on the data and does not have prejudice for/against it. Before the AC's jump and say that algorithms can be manipulated too, let me just say that they can be audited for soundness and logical mistakes/mathematically analized, etc. They are much more reliable than a human. There have been numerous instances when a correlation between factors was suspected, and data mining was used to prove/disprove the correlation or give a score to it.

So what's the right way to go here, you ask? I think that an image you can get from running an algorithm against the source is valid and can be considered an objective derivative of the original. Tools that allow to selectively retouch pictures should be out of the question. There are many techniques that can be applied to images to enhance them and that are wholly based on what's already there.
--

Re:Data Mining, "Airframe" and more.

[ebyrob]

A computer works on the data and does not have prejudice for/against it.

Maybe not, but programs often work better on one set of data than another... In this way they can be just as random as poor "human subjectivity". To think that a process is better just because it is wholly deterministic is silly. Actually, worse than silly, it is one of the original assumptions made by many modern scientists.

There have been numerous instances when a correlation between factors was suspected, and data mining was used to prove/disprove the correlation or give a score to it.

Interesting. "proving" something with data? Either that wasn't science or verificationism must not be dead after all... Regardless, different data mining techniques would each give a different "score" would they not? What should that difference be called if not a "bias"?

Re:Data Mining, "Airframe" and more.

[aWalrus]

programs often work better on one set of data than another... In this way they can be just as random as poor "human subjectivity".

No, they can't. An algorithm is not random. It is a process applied to data. This is just bullshit. Of course some algorithms will work better or worse on a particular set of data. If you can prove to a jury that the algorithm is not pulling an image out of its ass (which a human may), it is a valid point.

To think that a process is better just because it is wholly deterministic is silly

I'm not saying it is better in a general sense. I'm saying it is more trustworthy, repeatable and does not introduce external variables, which for this particular application, is better. An algorithm does not say "this line would look much better if I stretched it this way" whereas a human may. The algorithm does not care if the resulting image looks like shit.

Interesting. "proving" something with data? Either that wasn't science or verificationism must not be dead after all.

I think you're misusing the word, and Data Mining CAN prove that there exists a correlation or not in a set of data. For instance: a scientist may look at data for the foci of a disease in a country and see that there is roughly a correlation between locations of nuclear plants and the disease sprouts. He may then propose that there is a causal relationship. But after running the raw data through a data mining program, a much higher probability relationship may be found between temperature+water concentrations+disease (these three-tiered relationships are difficult for humans to find), showing that the former, seemingly correct deduction was just coincidental in nature.

Regardless, different data mining techniques would each give a different "score" would they not? What should that difference be called if not a "bias"?

Different data mining techniques are just different ways of analyzing a data set. They do not show statistically different results (they ALL agree on how much of each of the original data there is). Some of them are just better at finding certain kinds of patterns than others. It largely depends on the format of your raw data. If you have a point, make it. Processes don't think. They are not biased! is that clear?
--

Re:Data Mining, "Airframe" and more.

[ebyrob]

If you have a point, make it.

You seem to "trust" a deterministic algorithm to come up with a reliable and reasonable solution to some problem.

You seem to "distrust" a human being to come up with a reliable and reasonable solution to the same problem.

Does this mean in any problem set where a deterministic solution is viable and the answer "matters" the deterministic solution should always be chosen over a more human approach? If so, I'd have to wonder why humans still drive cars, fly airplanes and write algorithms.

Processes don't think. They are not biased! is that clear?

As a programmer, I assure you, I can write algorithms that show bias. Of course, you are correct that none of them think.

From www.m.w.com on bias:

3 a : BENT, TENDENCY b : an inclination of temperament or outlook; especially : a personal and sometimes unreasoned judgment : PREJUDICE c : an instance of such prejudice d (1) : deviation of the expected value of a statistical estimate from the quantity it estimates (2) : systematic error introduced into sampling or testing by selecting or encouraging one outcome or answer over others

Re:Data Mining, "Airframe" and more.

[aWalrus]

I'd have to wonder why humans still drive cars, fly airplanes and write algorithms.

Because computers still can't do it. If you think you can drive better than a computer you should also try to run a semaphore network or perform one million mathematical operations in less that a minute. Accept it: Creativity (and writing algorithms) may be the realm of people, but determinism is something a bunch of chips with software are much better at.

As a programmer, I assure you, I can write algorithms that show bias. Of course, you are correct that none of them think.

As another programmer, I assure you, that would greatly surprise me. Of course you can write a program that selects one answer over the other for arbitrary reasons! and you know what? That's exactly what it will do! It won't decide to go the other way just because it feels like it. You can even put in a random decision, and it will take a random decision! Don't you think an analysis of the algorithm would reveal the intended outcome? That was the original point of this discussion. I'm not advocating the rise of the machines. But I sure couldn't be trusted with an image and expected to come up with a totally unbiased, objective enhancement of it. Good or bad, an algorithm that doesn't try to replace the original content of the image and instead works with it will do a more reliable, repeatable job than I, or anyone else for that matter, can do (please note again that I'm not talking about a better enhancement, rather a more objective one! (which is what would be needed to bring this in front of a jury)
--

Re:Data Mining, "Airframe" and more.

[ebyrob]

Men and algoritms both tend to be judged by their fruits. This line from the article was probably the best argument for tossing out the technique at this point:

"There are no proficiency tests, there are no independent studies to say that this works or that it's reliable," Heyer said.

The same arguement would be made about absolutely any automated system for doing the same thing. Computers don't drive cars. Not because they *can't* drive cars, but because they can't do it better, or even as well, as humans.

Any algorithm written by a human being on a platform made by human beings ought to be just as suspect as a human being. If the basic algorithm were public knowledge and seperate open source implementations were created all with the same results on different platforms, then perhaps hard analysis could be brought against it. But a single implementation especially in poorly understood binary is subject to no such analysis.

This ACM classic on trusting trust does a great job of explaining one major loophole. How hard do you think it would be for a malicious coder to insert an algorithm that biased all results away from a specific set of prints?

As to software bias...
deviation of the expected value of a statistical estimate from the quantity it estimates

What is a data mining technique if not a synthetic statistical estimate? Bias in an alogorithm doesn't have to be about purposely favoring (or disfavoring) one outcome. Choose 10 completely different methods (or algorithms) for solving a particular problem (spam filtering would be a good example). Each solution to the problem will have different results even for the same input data. I would tend to call that difference a "bias". This bias would exist even with the most "trusted" algorithms and computing platforms.

But I sure couldn't be trusted with an image and expected to come up with a totally unbiased, objective enhancement of it.

Actually, if you were given a "rough print" you'd never seen before and had no idea what it was going to be matched against and had instructions only to "enhance" it, you might come up with something a lot less biased than you might think. As the technician stated himself, if he knew what he was trying to match against, it would be a lot harder to create an unbiased enhancement.

I'm saying [an algorithm] is more trustworthy, repeatable and does not introduce external variables

The algorithm itself *is* an external variable. The bigger and more complex it is, the more external variables it introduces. I'd argue about trustworthy, but I doubt we agree on the definitions of words like "trust" and "truth".

Resolution?

In the sense that an alogirthm in the right circumstances can be better understood than a human, and that a single algorithm could process more fingerprints than a single human, you may be correct that an algorithm could be more appropriate in something approaching an ideal situation. On different fingers, it should be quite plain that a human is vastly more capable than any set of algorithms yet contrived.

As to "enhancing" fingerprints in general. I have faith that our legal system can fully bung up either method. Human directed or computer controlled.

Re:Data Mining, "Airframe" and more.

[aWalrus]

You raise very good points, and it has been a pleasure to keep this thread going. Thanks for helping me dismember this dead horse :-) (I think we went past beating it about three posts ago). Hope to read more from you in the future. At this point, I think we've reached a good level of agreement. Thanks for the insight.
--

Encrypted cameras have another use..

[Myself]

Ever since I heard about Digita, the OS for digital cameras, I wanted to program something like this...

When the prosecutor subpoena'd photos from reporters, to identify people involved in the MSU Riots, the reporters obviously cried foul: Their independence is critical to the accurate gathering of news. If people see reporters as an arm of the law, coverage becomes skewed.

What if the camera could be loaded with a public key, and encrypt photos as they're taken, so that only the private key (back at the Editor's desk) could decrypt them? Perhaps keep a plaintext thumbnail in RAM for convenience, but make sure it vanishes if the card is removed or the camera's powered off.

It still doesn't prevent the judge from throwing the editor in jail for failing to turn over the private keys, but it adds a layer of complication, where the editor could simply lose the keys. Would that be destroying evidence? The encrypted photos still exist -- they're just unreadable.

Re:Tween the image and voila, it's anyones fingerp

[HiThere]

Fingerprinting is quite a reasonable approach. The problem is that they have quite loose standards as to what kind of match is sufficient. If they were only using it as a filter, to select who to investigate then the approach would be quite good. If they use it as primary evidence, then it's not so good at all (though still better than many choices).

N.B.: The same can be said about DNA evidence. Not only do the labs have a high error rate, they don't compare a sufficient number of match sites. (This MAY have been fixed since I last heard about it. The prices have dropped significantly.) If they only compare 10 sites, then that means that one out of (about) every two thousand people will match. (Assuming that they are checking for single nucleotide matchings, and even distribution of the traits. [Different presumptions would change this to around one out of every four thousand.) Now if you are screening suspects, this is pretty good. It will generally allow you to select one person to investigate more thoroughly. OTOH, if you are pulling people out of a database... there's likely to be a lot more than 4 thousand people in that database, and the culprit may not even be in there.

If they checked 20 sites... then the numbers start getting reasonable for primary evidence. If they check 30 sites, then it's a pretty good chance. (This is all assuming that they pick only sites that have high independant variability, of course.)

And there's no good reason not to use a pretty high number of sites. If you've got the DNA to test, you've probably got at least an entire chromosome.

Extremely slim chance

[Tenebrious1]

Any subtle change could incriminate the wrong person.

The chances are incredibly slim of that ever happening by accident. The chances that your fingerprint is mis-enhanced to say include a bifurcation, and that print matches someone in your neighborhood is almost zero. Maybe slightly higher that it matches someone in the world, but how many of those people would be fingerprinted and a possible suspect?

That's why fingerprints work because they have so many unique points, you only need to check a few of those points to get a positive match. Unless the enhancements are made deliberately, the chances of a misidentification are very, very small.

Re:Extremely slim chance

[dacarr]

That actually kind of was my point, on deliberate. Granted the odds are drastically against it, but think of it - from the perspective of a judge who may (not) know of technology (but has hopefully done their research), knowing that something has been altered, would you admit it as evidence?

Maybe this is me being a little paranoid about the encroachment of a day where technology is becoming more of a cradle-to-grave thing (and despite this I carry a Handspring Visor =) ). I could be wrong.

But apparently, not well enough...

[AzrealAO]

You said: Just because the person interviewed states he doesn't "create" fingerprints does not mean he cannot alter them, or "enhance" an existing print to more closely match a "wanted" print.

The Article said: "I don't think I could recreate a fingerprint," said Knoerlein, pointing out that he never sees the suspect's fingerprints.

similar to a polygraph?

[jmorse]

Polyraphs are generally inadmissible as evidence in criminal cases. I believe this is because polygraphs are not widely accepted as science. The same case could be made here, I imagine.

Inadmissible? Not so.

[MacAndrew]

Evidence gets "enhanced" all the time for clarity's sake. The court has wide discretion and the American preference is to let lots of evidence in provided it is not unfairly prejudicial, inflammatory, etc.

An example I recall was an effort to enhance the Rodney King beating video. It was shot at night and the operator didn't get the focus right for several seconds. One of the critical Q's I think was whether there were nightstick blows to the head, and by whom. I saw the before-and-after videos in a lecture by one of the prosecutors; unfortunately it just wasn't very helpful. But it was admissible.

I completely understand the concern about doctored or damaged evidence. Both happen, but not to the extent that all enhanced evidence should be excluded.

Where To Draw the Line

[suwain_2]

In my opinion, this isn't a case of "Should they be able to Photoshop fingerprints," but rather, "What type of modifications, and to what extent, are allowable?" I could draw entirely fictitious fingerprints in Photoshop; obviously, this would be an injustice if used as evidence. However, Photoshop is also great at salvaging bad photos -- with a bit of practice, I can 'feed' it a horrible underexposed image with a nasty green tone, and get a nearly-flawless image out. In my mind, this is no different than the adjustments people make in the darkroom -- if the forensics team used a film camera, and enhanced the contrast when developing them to make them look better, I bet no one would care in the least bit.

What I think might make more sense is if they were required to furnish both the originals and the 'post-Photoshop' version -- the defense could verify the work themselves if they wished. (If nothing else, they could probably find someone off the street to testify that, to him, he's not sure how they got the Photoshopped version, and that it doesn't look 'real'.

This might be enormously geeky, but if they really wanted to prove the integrity of the images, couldn't they bring in the computer and open up the Photoshop file, showing their work through the "History" window? You'd be able to show the result of "Open" (the original image), and then click on each thing you did to it to show the difference.

Trusted source, Open & repeatable operations

[xixax]

I was under the impression that some digital cameras already digitally signed images. This is what you would check in as evidence and what you need to take care over. And digital imagery in court is not new, my Prof. was using ehnanced Landsat MSS to argue irrigation breaches in court yeears ago.

Beyond faith in source data, I don't care what image processing operations are used because:
- I can trust that the source imaage is authentic
- The jury must be convinced that any image processing undertaken was reasonable
Hopefully the outcome is that basic enhancements (such as contrast stretching) is readily accepted, and more out-there techniques will need to be more persuasively argued. You would also hope tht the other side would attempt to replicate any operations that were suspect.

Xix.

Digital image certification

[xixax]

Regarding your link, nice idea, but that leaves the image unprotected until it is registered by someone I have never met. Either you trust the camera to sign the image, or you treat it like any other photographic evidence. There's no need to have a centralised repository signing my images when *I * can sign any digital file I produce.

Plus I can take additional non-digital precautions such as having wittnesses watch me put the media into a sealed envelope.

Of course getting people to safeguard their private keys is another matter entirely... :o)

Xix.

Re:Because it was the paperboys

[jedidiah]

OJ Simpson was just yet another rich man who bought himself justice. The only reason that anyone considers this remotely controversial is the fact that he was rich BLACK man.

Unlike the average man of his background who would be subjected to an overworked, underinterested public defense attorney, OJ was able to hire himself competent demagogues.

Above and beyond all of that, he LAPD and DA acted with careless arrogance. They chose not to avoid the appearance of impropriety and rightly got NAILED for it.

If they can't handle the HR details, why should anyone believe that they can handle the forensic details?

Re:Is Photoshopping a word???

[Dirtside]

Oh, please -- "photoshopping" is a perfectly cromulent word.

The real reason

[commodoresloat]

Come on. The real reason they should not be able to Photoshop pictures of fingerprints is that they should have to use the gimp instead.

loserish?

[commodoresloat]

How Loserish of you to think that ANYONE cares about proper verbiage.

Yeah. It's proper adjectives we crave.

An analogy with hard evidence

[Reziac]

Let me try:

In this case, DRM means "this file is evidence, and thou shalt not tamper with it, nor allow others to tamper with it". There is no issue of rights (or lack thereof), but rather of data integrity and provenance. This, IMO, is a rightful use of DRM.

Would YOU want to be on trial based on an image that had received an uncertain amount of twiddling by persons unknown? Of course not. You'd want to know that the image was correct, untampered, and that no one with an agenda (for OR against the defendant) had ever had access to it.

This is really no different than maintaining the integrity and provenance of physical evidence. Say you're arrested for drug trafficking, but in fact you only had a bag of sugar. Naturally, you'd want to be completely certain this very same bag of sugar is the one brought into evidence and presented in court, and you'd want to be equally sure that no naughty persons had dropped a spoonful of coke into the bag while no one was watching.

Think of the image file as the bag of sugar, and all should become clear.

You know, on CSI

[teamhasnoi]

they can make out a dude's face standing a mile away appear out of a grainy ATM video.

Just use those computers. Duh.

Re:Police-spec should be ok...

[Happy go Lucky]

.. There's no need for a law to force all users to have DRM in their cameras, but a police-spec requirement for legal preclusion of tampering would be useful.

How would you code something to prevent tampering? Admittedly, I'm a dumb-assed cop and no programmer, but I don't know how I could even define "tampering" in language that a compiler could understand.

Police already have legal requirements for speedo calibration in cruisers, and other standardization requirements.

Relatively few. Our policy (one city in Colorado-YMMV) on speedo calibration is to just let the factory do it. (On the Crown Victoria, the factory calibration is good forever so long as either the original tires stay on the car, or the replacements are the exact same size.)

Training requirements are theoretically standardized by state POST boards, but there's a LOT of variability there.

And about the only Federal standard we have (other than what the courts set) is the standards we meet in order to have NCIC access.

I would support police digital enhancement, just as long as it was strictly monitored and reproducible: that is, a secure source image must be used and provided so opposing technical experts could reproduce any enhancement when provided the methodology by prosecutors.

That's not that different from how we do it now. Any time I bring a photo into evidence, you can damn well bet that I can produce the negative for inspection/comparison on demand. Any image manipulation or enhancement needed, I can explain and walk the court through what I did and why I did it.

Unfortunately, without a negative, ultimately I'd have to ask the judge/jury to take my word for it. Our justice system is based on the fact that they're not supposed to do that. They're allowed to use their common sense (well, maybe he COULD have altered the negative with the super-secret machine brought down by the flying saucer aliens, but...) but they're also supposed to understand that "trust me" is just a polite way to say "fuck you."

It's that whole thing about an unaltered and unalterable negative which actually makes photography credible in court. Without it, we have a much harder time making our photos stick. I can start by testifying that the photo accurately represents the scene as I found it, but as long as there are people in this world with the ungodly stupidity to take Alan Dershowitz at face value, I have to plan on needing more than that.

As for the DRM

[Mac Degger]

Why not just use WORM media? Solves most of the problems, I'd say. Then all the auditing trail needed would be a secure IDtag on the media itself and sign in/out sheets.

Re:OT time

[HiThere]

Actually, the word I was deriving from was imminent. The concept that I mean is "bring nearer more quickly". (Admittedly, the more quickly part isn't included in the form...but it's a part of what I meant.)

It's a joke, in a rather black mood.

Just for fun.

[ebyrob]

Here are links to a couple articles I found on google searching first for "scientific naturalism", then for "philosophical naturalism".

Scientific Naturalism
Philosophical Naturalism

Re:Use a real verb

[RedWolves2]

Well you have posted countless times in this thread Anonymous Coward and none of them are even close to being funny.

Re:Use a real verb

[RedWolves2]

I do the guy is funnier then hell