Determining the age...
Clear all

Determining the age of person in pictures, esp. underage?

24 Posts
17 Users
Posts: 178
Estimable Member
Topic starter

In child porn cases, prosecutors request from our lab to examine the suspect computer and recover all pictures and video files with underage persons. So, we recover all indecent pictures and video files which we think might include an underage person and give those files to prosecutor, however, we can not be so sure about the ages of persons in pictures and this time prosecutor again asks us whether or not the persons in those files are really underage or are under a particular age, like under 15 or under 18.

As a principle, no computer forensic examiner can determine the age of a person in a picture or a video file just by looking the picture or the video. It may change from person to person and from race to race. No one can know for sure. Determining the age of person in real life is possible by some methods, but not so easy when it comes to do it on computer screen.

How do you deal with this situation?

Posted : 15/09/2010 4:57 pm
Posts: 650
Honorable Member

I don't do these kinds of cases any more, but when I did, you produced anything that might possibly be contraband and let the detective decide. I my old state (Qld, Australia) the detectives had the option to send them to the state censor for classification, and/or compare them against a database of known (identified victims) CP, or just pick stuff that fit the description, and let the jury decide.

Posted : 15/09/2010 5:22 pm
Posts: 283
Reputable Member

We get similar things here but it depends on the legal system.
In the UK, for example, we only suggest that the subjects in the pictures/videos are under age. The decision is made by the jury.

I suspect the best you can do is say that you have had previous experience with cases such as this and your opinion is that the person is such-and-such an age. You need to caveat this up like crazy though perhaps state that you are not qualified to answer questions of this nature and, if they want to get an expert's opinion on the age of the subjects, they should hire an expert paediatrician.

Is now the time to go on a long rant about how this is a ridiculous thing that digital forensic investigators shouldn't be asked to do? You wouldn't go to an electrician to fix a burst pipe. If you're feeling particularly uncooperative you can simply ask them to provide their own opinion on the age of the participants. Be warned I can not accept responsibility for any disciplinary hearings that might take place as a direct result of following my advice 😉

Posted : 15/09/2010 5:23 pm
Posts: 5133
Illustrious Member

There are (obviously COMPLETELY outside of the field of digtal forensics) sciences that do study relationship between looks and apparent age.

An anthropologists or a forensic anthropologist may be a better target for such a question.

Please note that the underlining of apparent, an expert may be able to formulate a more educated guess, but still a guess.

Expecially because the "underage" varies from country to country, and also with the actual crime involved, and on local Laws.

In Italy, just as an example, a 17 years kid (under-age when compared to 18) does not commit a crime if he has (obviously consensual) intercourses with a 15 years girl (over-age when compared to 14) BUT if he keeps on his PC some pictures of the act, or any image of the girl that could be considered underage pronography (subject < 18), as opposed to pedopornography ( subject pre-puberty) he is probably committing one.


Posted : 15/09/2010 5:52 pm
Posts: 3568
Famed Member

How do you deal with this situation?

I used to work with a former AFOSI analyst, and he and his team had received training from a pediatrician, which was then used as a basis for their determinations, particularly when going to court. At the time, he'd also used the contents of his military records to support his civilian work.

Posted : 15/09/2010 7:46 pm
Posts: 700
Honorable Member

As others noted, unless you have hash values which correspond to those of known child pornography, as a digital forensic examiner, determining the age of the subject is likely outside your area of expertise.

And having worked in some CP cases, I can confirm that there are filesets that include women who are of legal age but appear to be much younger, either naturally, or through digital alteration. And while it is true that there are certain morphological features that can be used to estimate age, the accuracy of these estimates vary with the degree of precision required, i.e., it is much easier to correctly estimate the age of a person within 10 years than it is to correctly estimate their age within two.

That having been said, it is much easier to estimate the age of young children than it is to accurately estimate the age of adults because children go through so many morphological changes in the first two decades of life. Still, such predictions can be rift with problems as in the case of the man who was arrested for CP after videos of Lupe Fuentes was found in his luggage. Two experts, including a pediatrician, testified that she was under 18 in the videos but the subject, herself, testified that she was not and produced proof of her age.

Part of the problem is simply natural variations in the rate at which people mature and develop the typical characteristics of adults. Part of the problem is that external factors, such as nutritional status and participation in athletics can alter the appearance of many of the markers of "adulthood". Consider how difficult it was to establish, with certainty, the age of Chinese Olympic gymnasts without the actual birth records.

Even forensic pathologists have a hard time predicting exact age, especially as the person approaches adulthood.

The best machine guessing systems are only roughly 50-80% right and that is within 10-2 years.

All of which to say that it can be risky to rely on such opinions, especially if one is dealing with late adolescents. Certainly, I would not want to venture such an opinion.

Posted : 15/09/2010 9:35 pm
Posts: 5133
Illustrious Member

Just for the record, an article summing up what happened

And here is a copy of the affidavit of the ICE "special agent" about the re-known Lupe Fuentes mishap
and the notice of intent for the "expert witness"


Posted : 15/09/2010 9:59 pm
Posts: 700
Honorable Member

Just for the record…
…the notice of intent for the "expert witness"

The problem lies in the fact that many true markers of maturity can only be detected through physical examination and testing. What experts typically use for CP images is the Tanner scale which was not designed as a predictor of age but, instead, as a means of assessing individual children with respect to their expected stages of development at various ages.

In pediatrics, this is no different than measurement of height, weight and head circumference which are not used to predict age but to where, on various statictical curves, the individual lies with respect to their peers.

And while the aforementioned case is, perhaps, the most famous (in part because the film's star actually showed up in court to exonerate the suspect), it is certainly not the only case in which charges were dropped or thrown out after suspected CP was identified as not being CP (e.g., Melissa-Ashley).

These are, however, good examples of the dangers of basing an expert opinion on unscientific methods.

Posted : 15/09/2010 10:30 pm
Posts: 45
Trusted Member

I have never had to do a case of this nature but I did see on the news one day that Dr. Cyril Wecht (a noted medical examiner who has been called upon many, many times) was brought in to examine images to determine the age of a 'victim' in a CP case.

I met him about 12 years ago when his son David was running for office as a judge (not sure what level) and I met his other son Ben when I attended a Computer Forensics conference in Pittsburgh last fall. Duquesne University has named their forensic science department after him.


BTW-He also holds a J.D., as well as an M.D.

Posted : 15/09/2010 10:55 pm
Posts: 5133
Illustrious Member

These are, however, good examples of the dangers of basing an expert opinion on unscientific methods.

Yep, it is also interesting as I see it, the actual wording of the notice of intent

Summary/Bases/Reasons of Testimony – Based on his numerous years of experience and training in treating children, Dr. Jaunarena will provide expert testimony regarding, but not limited to, the age of the minors depicted in the images in the labels of the DVDs and in the video images found in the DVDs seized in relation to this case. During his testimony he will explain from viewing the images in question the bodily features he considers in making his determination, such as the face of the minor, the breast area, the genital area to include the existence or non-existence of pubic hair, the height of the minor, among other factors to be considered to establish the approximate age range of the minor depicted and to establish that the relevant images depict minors under the age of eighteen.

As often happens I am picky wink , but using instead of something like

of individuals that are thought as being minor

using the "direct"

of the minors

it seems to me like treating the hypothesys as if it were the actual conclusion.

I.e. logically an "expert witness" should first decide whether (in his/her opinion) any minor is depicted, and then give his/her opinion on the apparent age of those.

And as well, whatever may (or has actually) come off such a testimony is an opinion, and in my book an opinion has never established anything, actual evidence or the actual verdict may establish something.

The "explanatory memorandum" here
sums up nicely how what an "expert witness" normally brings an opinion to the court, educated and backed up all you want, but still an opinion.

AFAIK, in the US also the "Daubert standard"
is replacing in most states the "Frye standard"
and it's "general acceptance" with a more "strict"

Scientific knowledge = scientific method/methodology



Posted : 15/09/2010 11:31 pm
Page 1 / 3