Although, we haven’t yet been challenged in court, I would suspect that if called to do so that we would at that point expect an Adobe expert to be called to provide such details to please the court. I base this on other types of evidence we provide such as testing of drug samples using the Beckton Dickson NIK testing. When challenged in this manner, a representative from Beckton Dickson provides such information. Now, this is in Canada. In other jurisdictions, things many vary considerably.
Dave
Thank you Tony, Not only do i need a technical explanation but I need to be able to explain the process used to a jury and judge. I am definately not looking for source code information, but a way of understanding the process in a scientific way. I would then be better prepared to explain it to a jury in terms that they could understand. In past trials I have used the "traditional darkroom counterpart" of the tools used to explain the application. For example (dodge and burn, contrast, brightness, and so on).
Dave Hamer, When the technology of forensic digital image enhancement was recently challenged in a "Frye Hearing", I contacted Adobe to request a expert for testimony. I was told that the legal division at Adobe would not allow anyone to testify for either side in any criminal case.
David,
Then your quest would largely depend on the actual techniques used to enhance the print. Again, more "traditional" tools can be explained reasonably easily, again using things like "an unsharp mask was performed which finds the difference between light and dark edges". Once they need to know HOW that’s done, there’s plenty of documentation on the process without having to go into a specific algorithm used – at least I would think.
When you talk about resizing an image using Bicubic resampling, you can obviously speak to the issues surrounding the removal of pixels to accomplish a smaller size, but if I recall correctly, Adobe uses a proprietary algorithm to re-sample. If that’s called into question, then almost certainly someone from Adobe would have to be issued a subpoena.
The "fit and finish" of it all is, I’m not 100% sure that this forum is going to be the best place for your answers. Rather, I think that if you’ve documented the steps performed, you would have to list those steps out one by one, then use information available in the help file for understanding exactly what a particular step does. Then, do a google search on some of those techniques to see the theory behind it.
At some point, a good attorney with a good expert witness (for the other side) can ALWAYS challenge a particular technique and the algorithm used (for example in using a blend mode). Let’s face it, the fact that you are shifting pixels, removing them, darkening them, is changing the image. At that point, you would need someone who can discuss the issues surround the algorithm, what it does and what it does not do.
Forensics with digital imaging can be tough since there is always the argument that you are fundamentally changing the image. However, digital enhancement of video footage to, say, identify a car at a crime scene from a video still, goes reasonably unchallenged.
I’m not sure I have provided much more information than you had when you started your quest, but again, I’m not sure this forum will be able to help much.
<shrug>
Peace,
Tony
Could this be a case where obtaining the ACE title (Adobe Certified Expert) on Photoshop) be of some help?
I mean, an official paper whereby the producer of the software certifies that a given individual officially ‘masters’ the tool should give his manipulations more ‘reability’… Or shouldn’t it?
David Knoerlein
Thanx for that info. We will have to keep that in mind in case we ever face the situation. I suspect, it will only be a matter of time. In speaking with a counterpart several months ago now, he advised that we simply have to explain in laymans terms what enhancements we had employed and the reason we felt them necessary.
Dave
In a way Gustavo, perhaps. But at the end of the day. It doesn’t matter who you are, you have to be able to explain it if you are challenged by an equally knowledgable person on the other side of the argument. From what I’ve seen of ACE, it doesn’t give you the scientific theory as much as practical application. And even if it did, you have to be able to defend your position.
The mere fact that you are called into court to explain what you did, you are considered an Expert Witness…
David,
If you have to explain it in laymans terms, then the online manual and a good google search will be your best ally.
Tony,
What you say makes sense. Well, another idea down the drain…
I’d think that the court would rather have a Senior Product Manager for PhotoShop give an explanation of how something works, rather than a user, no matter how qualified.
I can certainly understand Adobe wanting to stay out of the courtroom. If called into question, their intellectual property would become public knowledge.
I’d also think that Adobe’s reputation can speak for itself when called into question. For example, if they were to ask what the chances are that extra pixels can appear, I’d think (and I’m far from a lawyer) that Adobe’s reputation as being the leader in advanced digital and print photography would mean that absolute clarity to the highest degree must be maintained at all time.
If it’s SOP for you to use PS to scan fingerprints, that means that PS was certified as the best product for the job.
Hey everyone! This is a way to integrate PS as an OEM product. To have it being the software running a fingerprint station. I know, OT.
I’d think that the court would rather have a Senior Product Manager for PhotoShop give an explanation of how something works, rather than a user, no matter how qualified.
Cheese, it depends on the point being discussed. The user is an expert witness in that he used his tools to arrive at a conclusion that brought them to court. He is to testify, explain, and defend what tools he used and why. The why part is his domain, not the Senior Product Managers.
If it were called into question whether the technique was sound, this expert witness would site literature, and possibly require someone from Adobe to testify.
But the fact is, what is really at issue is: What was done to the evidence (fingerprint, urine specimen, etc.) and why? Once that’s established the "other side" can question the validity of the choices made by the expert witness, in which case he/she must defend it.
So except as a reference to prove why a particular technique was scientifically sound in arriving at a conclusion, it is really the expert witness, in this case David, to defend the procedures that lead him to a conclusion.
Also bear in mind that the whole point is Cross Examination; an Adobe Senior Product Manager, may, in some cases, help the other side unwittingly.
Peace,
Tony
David,
OT or useful, you decide on it.
First, I do not work for, own stock in, or as far as I’m aware of, don’t even know anyone in ANY software company.
I recently looked at FocusMagic (NOT giving my impression of the product OR answering "Did you or did you not purchase said product?")
They had an explanation of what their product "looked for" and what it did when it found that condition. IF this meets your needs of what and how, then you might want to try the limited usage download for sharpening.
Odd when you have to use software because of the provided information on how it works, but in your case…
www.focusmagic.com
Bill
I’m not a lawyer, but since Photoshop is not marketed as a forensics tool, I wouldn’t think they could reasonably be expected to divulge trade secrets in a criminal trial. If a program were marketed for forensic use, I surmise there would be such an obligation.
I’d think that the burden is on the forensic technologist/scientist to explain why they chose Photoshop, and a specific tool in Photoshop, to do a particular enhancement. If you don’t understand how the tool works, hence what artifacts it might introduce, should you be using it on criminal evidence?
Terry
Thank you everyone for your responces to this important topic. Let me tell you that during the Frye Hearing in the case that I was involved in (St. of Fl. vs. Victor Reyes) the defence challenged the use of this technology in as a forensic application. I have developed and implemented a very detailed SOP for the use of digital imaging in our agency that covers the guildlines for enhancement. These (general guidelines) follow the recomendation issued bt SWIGT, the Scientific Working Group for Imaging Technologies. To address the use of Adobe Photoshop as a forensic tool the state called myself and two other experts to testify. Together we were able to explain to the court that the tools used in this case did not alter the image. The case gained national attention when "60 minutes" aired a segment entitled "The Hidden Clue", which focussed on the defense’s arguments and the forensic communities answers to those arguments. The issue of whether an image is altered or not during the process was discussed. I have been doing digital image enhancement for over 6 years now and forensic photography for 16 years. In the end when the dust clears, it all comes down to the integrity of the individual performing the enhancement.
PS. We won the hearing and the enhanced image was excepted into trail.
My penny.
Would the court accept a live demonstration on site? E.g., capture an image in the court room, and process it with PS. Then submit the result to the court. Repeat as many times as wished.
Is this a foolish suggestion?
Is this a foolish suggestion?
Not foolish, but not helpful, and sometimes damaging. Couple of points.
Do you want the jury to see "magic"? Probably not. That isn’t to say that there isn’t a benefit from before, intermediate and after images to show a transition, but watching a clocks hands move does nothing, really, to explain how time is kept. If you get my meaning.
There has also been work done on digital simulations in the form of video – the prosecution proposes what happened at the crime scene, and plays a simulation with either computer generated characters or actors.
This can play either way with a jury, since it is being acted, it’s obviously not real.
The defense contends that it contaminates the jury’s mind because even though it IS a simulation, "seeing is believing".
With that in mind, the defense could make a similar case with the judge that showing photoshop, for example, unfairly biases the jury and thus there is questionable utility for either side in a software demonstration.
Better to move along the path that David has by getting Photoshop (or any other program) recognized as a forensic tool that produces submissable evidence.
Peace,
Tony
It’s also very easy to show how to put my head on Pam Anderson’s body. And isn’t that what PS was designed to do in the first place?
Just think how unpopular internet dating would be if we didn’t have PS.
You have a very sick mind Cheesefood – I like it.
But your point is well taken – the value of image manipulation is a pendulum.
Sorry, I was focused on fingerprints.
I know Sam, but the same priciples apply there. If you can alter an image, you can do so for good or evil. Demonstrating it just shows how easy it can be.
Typically, in my line of work. We have been asked to explain it in what I refer to as laymans terms. No one is expecting us to be able to go into a lot of specific details because the Judge and certainly the members of the jury would all think we were speaking geek and would quickly be lost. If the challenge is between the defence and the pocecutor then usually the experts from the company who supply the products are forced to attend by subpoena. When you receive a subpoena, there is not much you can do. If you explain to the Judge that you want to protect your source, then often they consent by way of a voire dire.
Dave
Sam, Thats exactly what was done in the courtroom. I demonstrated each step of the process and provide the court with a detailed report of the enhancement history. Again I must say that the court excepted the process as a valid forensic tool. What I’m looking for is assistance in where I can find logical explanations of what exactly happens to the "Pixels" as each enhancement is applied.
David:
I would show a physical example of how the enhancements detailed the image, eg. photo 1: original fingerprint scan, photo 2: level adjustment etc. this would show the jury in real terms how you arrived at the final image that was used to link the suspect to the crime.
If the jury has a question about how a particular enhancement step (eg: "I noticed certain changes in the fingerprint during this enhancement"), you can focus on that concern, rather than other steps that do not give rise to any concern.
ID. Awe
Agreed. In regular film photography, one has to be prepared to show the negative. With no negative available to digital, one has to be prepared to show the original image and the then the enhancement and be able to describe how it was enhanced and why. This is the way it is being presently accepted in Canada, at least according to my RCMP Ident specialist.
Dave
A word about "Alterations". In accordance with the SWIGT guidelines and my own SOP’s, only tools that do not add or remove pixels will be utilized to enhance images of evidence. To ensure security and to manage the images I use a (Forensic Image Tracking Software) that applies an encryption code to each original image at the time it was acquired into the system. A dupelicate of the original (also encrypted) is launched into Photoshop, and afetr being enhanced is returned to the tracking software database. The encryption code is then verified. If a single pixel has been added or deleted the code will be different, thereby displayed as a corupt images. This and many other security measures have become policy for all images of evidentual value that are captured ditally.
Excuse my curiosity,
What does it mean here ‘adding or deleting’ pixels? I mean, does that include changing its values beyond a given range?
Thanks.
Gustavo, increasing or decreasing the pixel count for that image would be an alteration. Adusting the values is concidered a enahncement. I explain it this way in court. When I do enhancement to an image its very simular to adjusting an image on a television set. I adjust brightness, hue, even volume to give me what I beleive is the best image for viewing. I do not change the contents of the image on being broadcasted to my TV, I just make it more clear and easier to see details. This is just like the enhancement of digital images, which is a (SUBJECTIVE Process)and up to the analsyt to decide what is used and how much of a change to make.
No worries about the spelling and grammar, David. It’s clear those are just hasty mistakes, given your grasp of the subject and depths to which you’ve gone to share your knowledge.
You can click the "Edit" text link just below the body of your message for up to 30 minutes to make corrections, addendums or to clarify.
Just for gits and shiggles, please consult Seymour Tipps <
http://www.adobeforums.com/Images/help/edithelp.html>.
Some operations can unwittingly remove a pixel, no?
If you adjust the values to far, the image will pixelize, thereby removing the pixels from that channel.
Is this correct? Does your encryption software catch that kind of dropout?
On second thought, the pixel remains; it’s value is zero. Hmmm.
David,
Thanks. I asked because I think a lawyer could rightfully argument that changing the values of the pixels may be even more important than adding or substracting one.
That is: If a digital picture is basicly a big spreadsheet of data, the content in each cell and how they interrelate is as important as how many cells there are.
Perhaps it would be better not to get into unneeded details, as it was mentioned before 😉
Most of these issues are pointed to by the book Rising Sun by Michael Crichton. Again, image manipulation in any fashion can be used for good or evil, whether intentional or unwittingly.
I think David has a most excellent handle on the issues. One of the main issues is his documentation of his methods. Those demonstrate the rationale and will be hard to argue with. They DO change the image, but it’s not difficult to establish that they are techniques that enhance data already in the image and removes "noise", similar to the techniques used with analog video stills.
So A)establish the protocols as sound techniques and B)prove that you followed those protocols. That’s really what is at issue.
