Schneier on Security
A blog covering security and security technology.
« Spam as a Business |
| Two Good Rants »
June 10, 2011
New Airport Scanning Technology
Iscon's patented, thermo-conductive technology combines infrared (IR) and heat transfer, for high-resolution imaging without using any radiation. The core of this is state of the art imaging which detects and processes a break in the established thermal balance between the clothes and a hidden object. The IR camera detects the heat radiating from even a tiny object, producing a dark/light shape. It is irrelevant how long an object is concealed under clothing as a new temperature imprint is created every time it is scanned. Using IR, the rays don't penetrate beyond the clothing so there are no privacy issues.
EDITED TO ADD (6/14): Another article.
I know no details.
Posted on June 10, 2011 at 6:14 AM
• 41 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
"I know no details."
Hmm well at first glance this,
"The core of this is state of the art imaging which detects and processes a break in the established thermal balance between the clothes and a hidden object."
Sounds wrong, I suspect it actually detects a break in the established thermal balance between the clothes and the body caused by a hidden object.
That is the human body within reason can be considered over small areas a "uniform radiator of IR energy" likewise most clothes can be considered over the same area "a uniform absorber / reradiator transmission line of IR energy".
Any object in between changes the absorbtion / transmission over an area, so if you have a sufficiently sensitive IR camera and edge detecting / enhancing software the ghost outline of objects inside or outside the clothes can be seen.
However before I do any further guessing I'll go wrangle the patent (which I know Bruce recomends against for very good reasons ;)
I am left with a "hmm."
Clive - does anything show what its success rates are?
Surely the only way this can work is if there is a significant difference between the body temp and the "object" temp. Wont most materials eventually balance out?
@GreenSquirrel "Wont most materials eventually balance out?"
I'm thinking that the outward face of the hidden object wouldn't be the same temp as clothing. The objects inner side would absorb body heat. But it wouldn't cook through or absorb from the sides, much. And it would/could radiate at a different rate.
Perhaps a thin small object, or a large large person. Would it detect something in the folds of flesh?
To my list add "This blog needs a lab. Infinite budget would be nice."
Your link appears broken or truncated.
"Won't most materials eventually balance out?"
Only if the ambient temprature is the same as the human body and "heat stops flowing".
I've not got the patent yet, but there is another artical on it,
And this definatly suggests it's image enhancment as well, as it's described as,
"Iscon Video Imaging said TSA is testing its patented Thermal Boosted Infrared Detection System"
It sounds like a cross between edge detection and AGC gating. I worked on edge detection in images back in the 1980's with body scanners to enhance things like feignt line images of hairline cracks etc. You can also do similar for soft tissue swelling etc simply by using highly nonlinear contrast adjustment to enhance very minor "brightness differences".
no details = no information = ??
If it's true, it's just a slight refinement of existing & widespread IR camera/video technology. Kinda like going from an 8 megapixel personal photo camera to a 16 megapixel version, with better software.
Practical application would be very difficult (airport passenger screening) due to the vast array of human clothing & carried items that can easily mask IR signatures, or confuse IR sensors. Low signal to noise ratio.
Simple vegetation and tree cover hides human IR signatures against the most sophisticated police/military airborne IR video cameras. Savvy bad-guys would have no trouble hiding weapons/contraband with IR absorbent cloth or packaging.
The TSA wants a new technology to replace its controversial X-ray scanners. It must be high-tech to reassure the terrified, completely harmless so that people can't opt out of it, and preferably expensive because-- well, because.
Who cares if it works?
They will foolishly believe it's more effective because it's "newer" and therefore has "better technology" regardless of whether or not it is. If they'll believe that, I'd much rather them go with this device that, while as ineffective, is less dangerous and intrusive.
I'll venture a guess, so as not to appear as if I actually know anything about it :-|
The IR detectors in question may have some spectral discrimination, rather than being simply one broad-band filter. By fitting multi-filter data, they get a better spatially-resolved temperature map than would be possible simply by assuming a black-body intensity-temperature relation. If they can actually spatially resolve a degree Celsius or so, they could almost certainly resolve density inhomogeneities, which as pointed out above radiate and cool inhomogeneously, and therefore should show up on sufficiently high-resolution temperature maps.
I will try to clarify my previous question with an example.
Is this system able to detect a handgun in a leather holster, in my armpit, after I have worn it (and a coat / jacket) for most of the day? Will there still be sufficient heat difference?
I've found three patents for Iscon Video Systems,
1, 2010016511 : SYSTEMS AND METHODS FOR CONCEALED OBJECT DETECTION.
2, 20090294667 : METHODS AND SYSTEMS FOR DETECTING CONCEALED OBJECTS.
3, 20090110232 : METHODS AND SYSTEMS FOR DETECTING CONCEALED OBJECTS.
I'm guessing the first one is the patent in question.
It's abstract reads,
In an exemplary embodiment, in the infrared range, in practicing the method of these teachings, three infrared images of a body having concealed objects are acquired... ...The responses of the detector to each of the three images are utilized to provide RGB signals to a display device... ...The color image displayed in the display device allows identifying the concealed object...
Makes it sound like simple spectral upshifting to give a "false colour" representation semi optomised for the human eye-brain response.
In otherwords on the simple assumption lower tempratures are going to be read shifted and hotter tempratures blue shifted. When the ambiant temprature is aproximatly 10deg C lower than body temprature objects of low heat transmission (denser object) will apear red in comparison to high transmission (naked skin) or normalised clothes transmission being green.
With respect to your gun in it's holster simple answer is yes.
If you think about it if you have an object that is a poor conductor of heat between you and your shirt, where the shirt touches you it will give out more energy (ie loses heat) than where the shirt is in contact with the poor conductor. So the shirt will be warmer where skin contact is than where the poor conductor is.
But as I said this only works if there is a sufficient temprature differential between the human and the air they are in. So I suspect the aircon bill will be high with the definate "chill in the air feeling" not being due to "TSA junk grope fear" (AKA "Black glove and tourch syndrome").
A thought occurs, this site may have "prior art" on it I vaguly remember discussing spectral spatial discrimination detection for this sort of thing in regard to already developed IR technology for satellites doing multispectral (ie 256 or more frequency bands) imaging.
The patent shows an IR load under which the subject is placed. Probably a heat lamp or cold draft. It seems the goal is to take a reference image, put on the IR heat and take another image. You're really trying to image the heat capacity of the object through thermal inertia. Presumably folks wouldn't mind being irradiated with IR photons as much as X-ray photons. With a short enough IR pulse, you might not notice you're being irradiated. Oh, but then you don't feel the minuscule radiation from the current naked scanners.
What's the price tag? That's the real question. Given that none of these devices even claim to be effective against body-cavity concealment, could we just pick the cheapest noneffective device?
"Using IR, the rays don't penetrate beyond the clothing so there are no privacy issues."
I'm not so sure about that. There was a bit of a kerfuffle a few years ago when people discovered that the darker patches of skin radiate and/or reflect IR differently than the lighter patches, and that cheap camcorders were fairly sensitive in that part of the spectrum. Some fabrics are not as opaque to IR as they are to visible light.
"It is irrelevant how long an object is concealed under clothing as a new temperature imprint is created every time it is scanned."
What is that supposed to mean? The first half of the sentence seems to imply some reason why an object which has had a long time to be body-warmed would still be detected, but the second half looks like it's about something else entirely.
"high-resolution imaging without using any radiation"
Well, it uses radiation -- even if it's not "radiation."
"So the shirt will be warmer where skin contact is than where the poor conductor is."
Now, my next question is will this be enough if you've been carrying it for time. While I have never tried to thermally image it, when I have spent a lot of time with a handgun in a holster, by the end of the day, the weapon "feels" pretty close to body temperature, so the question is probably - how much temperature difference is there and how much would this device detect?
I notice on the pictures on the webpage, there are big huge blobs where the suspect equipment is - I assume this is a result of staging objects with massive differences rather than a live test.
The problem I see with this is that it wouldn't be able to tell the difference between like shaped objects (e.g. a piece of cloth and a thin sheet of explosives) OTOH that still a win as it makes the effort to get said stuff past security that much harder and may well catch other components (e.g. a reasonably effective detonator)
I think the bulk of the IR radiation from a human blackbody (racist!) comes from inside, where temperatures are significantly higher than the skin.
So even if a gun was warmed to skin temperature, it would still produce significantly less IR radiation.
Still, you have to wonder - what if the gun were somehow warmed to internal body temperatures?
I plan on submitting a patent for a device to reduce the fear of terrorism. It is a plastic porticulus through which passengers are required to walk. An IR sensor detects the presence of a passenger walking through. When a passenger walks through, there is a 1 in 10 chance that a siren will be emitted so the passenger can be pulled aside and run through a different machine (also patent pending). Every time a person walks through, a heat lamp and a clicking noise are activated. At the side of the device is a standing console comporting between 100 buttons, knobs, blinking LEDs and switches all labeled with names such as "Polarity" "Sensitivity" etc. The machine's operator must be extremely rude and self-important.
"so the question is probably how much temperature difference is there and how much would this device detect?"
Realisticaly we can measure better than 0.01of a degree C without to much difficulty.
Now as somebody above has mentioned the patent apparently has a thermal load. Now you can measure the noise floor in an amplifier by using a hot and cold resistive noise source (ie fifty ohm at room temp, fifty ohm in liquid nitrogen and a coax switch to switch between the two) which enables us to measure quite accuratly to within a few dB of the true thermal noise floor.
With a little care the same switching method can be used to artificialy create a surface temprature change, thus you can measure "thermal lag".
I won't go into the details but a certain mathemitician (fourier) came up with a new branch of mathmatics to measure thermal lag in various materials.
Now the question arises what effect does a jacket with high insulative properties have. As we know from the space shuttle tiles we can make materials that realy do not transmit heat that well as well as metal tubes with an apropriate liquid vapour transition that can conduct heat faster than the equivalent sized copper pipe.
Thus with not two much effort it should be possible to build "heat shunts" around voids etc.
But much simpler as has been said before make a leatherette effect coat out of a suitably maluable plastic material lined with a uniform layer of plastic or other explosive. You can have decorative welted seams that are actually RDX fuse etc... And your Ipod or other backlit display to generate a high voltage charge that you store on a capacitor to fire a home made thinwire and chlorate/fulminat/permanganate detinator (or simply use the scrapings from red match heads you have disolved and reformed via a miniture flashlight bulb filiment then coated with another primary explosive.
It would not take to much effort to test the leatherette composition using non explosive materials with similar thermal properties.
Looking at IsCon's patent (7664324) it looks as if the person being examined is placed inside an enclosed booth and warm air is blown on them from an array of blowers which rotate around so all parts of the body warmed up. The IR cameras then look for differences in thermal conductance. Presumably an object under the clothes that are not skin would show up as being a bit (
Are you the same person who patented those text-pagers that the police use with dim-witted criminals, which can detect lies when the criminal has his finger held against the screen?
"Realisticaly we can measure better than 0.01of a degree C without to much difficulty."
Excellent - is that the sensitivity of this device? Does it do that at a distance or is this in "ideal" circumstances?
@PrometheeFeu = FTW.
I can see it now, the market is ripe for a new line of IR absorbent holsters! Passive models that match the radiance of a normal human body or active models that match the wearer exactly. The next must-have accessory. Or maybe even IR reflective shirts that have an inner layer that absorbs the normal body IR signature. Now where's that patent application?
The thermal measurment at slightly less precision range at -40 +200 and can be done opticaly, however you have to know the angular spread of the optics.
You sometimes see them used for measuring difficult to get at components.
However it's not the absolute precision you are interested in but the differential sensitivity.
No "radiation?" Check!
Less nudeiness? Check!
No more or less ineffective? Check!
Still eye-wateringly expensive? Check!
Still locking the door to the flight deck? Check!
Still flying with passengers prone to beat the crap out of you for acting terroristy? Check!
What's not to like, unless you're a former DHS boss shilling for a company offering a "some radiation with nudeiness" solution?
Got anything to detect several ounces of C-4 or Semtex jammed up someone's butt? How about a detonator jammed up another passenger's butt?
Right, prostate palpitation...Definitely where TSA will be going next. :-)
End result: Oops, blew the guy up at the check station...along with ten other civilians...Stupid terrorist had the detonator up there along with the C-4...
"Abdominal x-ray, pelvic exams, prostate palpations...easy, peasy We've been working on this issue over at LinkedIn groups."
Fair enough PR is supposed to be helpful
"Decided it was a branding problem"
I am not going for the red hot poker treatment, you'ld not be able to sit for the whole flight...
And as BF Skinner noted the other day, some US citizens have a bit more to keep covered up than others, so me thinks that with branding subsiquently there would not be enough ice on the flight to go around...
Here's how to foil the system: Wear two layers of clothing. The undershirt should be a close but not snug fit. The overshirt should be loose. Both shirts should be made of a relatively thick fabric that tends to scatter IR radiation instead of transmit it. (Similar to fabrics designed to reduce transmission of UV radiation and prevent sunburns.) Objects that alter IR transmission from body skin will be difficult to detect because the IR scattering of the shirts and the air-filled gap between overshirt and undershirt will eliminate distinct "edges" and thereby foil the sensor/software combination.
Why not force all TSA employees to work naked, if they are going to come up with new ways of imaging peoples bodies directly or indirectly via thermal imaging? Since when did nudists take over government agencies?
There is a claim that it won't image through clothing, but it is practical to configure a good thermal imager to resolve body shapes and body parts including genitalia through clothing, since these radiate heat.
I wish there existed a separate passenger flight system with the same security screening as for rail travel. I would then have a much more desirable option, since I value my liberty and dignity over my fear. Heck, they can even install remote self-destructive devices on these planes, if they are worried about them being highjacked.
This sounds like a expensive IR imager with some fancy edge detection software. Just one more expensive fancy toy to molest the public with.
Our (rocky Flats Gear) x-ray/mm-wave airport scanner protective radiation shielding garments would perhaps be a 0.5-2 C cooler than the skin. This and other items would be difficult to detect with such a low S/N with a variety of IR opaque clothing, it would be easy to defeat. Still has the legal search issues, false positives lack of probable cause for a "pat-down".
I think you are way over complicating the whole system. If you want to see behind a dominant reflection (or emission) than with a microscopic you use a technique called co-focal imaging.
The basic method applies equally to any area of optics so it will work for IR body imaging.
The secret of the technique is that with a large aperture you have a very short depth of field. what this means is that only objects in a small window are in focus. If you locate a pin hole at the focus point than everything out-of-focus gets removed and only the information at the focus gets through the pin-hole.
This technique can get between 20dB and 40dB rejection of unwanted signal.
The trick is maintaining a very accurate focus and tracking the object as it moves.
I've never tried this but there is no reason that a synthetic aperture optical phased array system could not be created, this would give you a huge effective aperture.
Additionally there are spectroscopic imaging techniques that will get you about another 20 to 30dB, although I'm not exactly sure of what materials are needed to do this at 8000nm to 10000nm detection wavelengths.
@Lisa: Heck, they can even install remote self-destructive devices on these planes
All planes in any Western country already come with such devices. They're called heat-seeking ground-to-air or air-to-air missiles and are brought to proximity of the plane which needs to be blown up right before the destruction commences (which is significantly better than having a destructive device permanently bolted-on: less risk of accidental discharge).
termal imagers for under-the-clothes detection?
A truly stupid idea. Make a shirt from a space blanket; wrap it into some electro-conductive material which heats to 33 or so degrees (approximating average body temperature at the torso skin). Make it so that seams coincide with the seams of the outer clothing, so they won't be noticeable. Now you can hide anything under this IR camouflage.
This probably uses active IR thermography, meaning they heat the sample and measure the response. Generally the amount of IR radiation will be much greater than that of the surroundings.
If enough energy is used such that it transmits through a layer of clothing and part of it reflects back it could be possible. Also, many plastics (namely LDPE and HDPE) are translucent for low heat fluxes so a plastic jacket may not offer as much protection as you would imagine.
Defeating this technology should be pretty easy, all you need to do is change the emissivity of the area where you're concealing something such that they can't determine what the object is.
"high-resolution imaging without using any radiation"
Well, it uses radiation -- even if it's not "radiation."
True, I suppose what they mean is that they don't 'add' any radiation, the imaging should be done completely passively, using the human bodies to 'illuminate' the view.
Oh, I'm 100% sure a stiffy would show up on this thing. I still remember the IR video in high-school health class of an erection. No doubt something sensitive enough to pick up the differences between skin & metal would be able to see skin vs. excited skin.
Schneier.com is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc.