(Fwd) British Study Claims That Photo Credit Cards Don't Work

jim bell jimbell at pacifier.com
Fri Apr 12 14:28:46 PDT 1996


At 08:42 AM 4/11/96 -0400, Clay Olbon II wrote:
>At 11:10 AM 4/9/96, jim bell wrote:
>>At 07:57 AM 4/9/96 -0500, Mike McNally wrote:

>>>There are supposedly some new techniques that look at the infrared
>>>signature of your face (like, I guess, distribution & position of
>>>hot & cold spots), and that's less likely to be fooled by facial
>>>hair and other superficial disguises.  It's probably a fairly simple
>>>technology, and could be applied to the credit card ID problem.
>>
>>I think this is based on looking at your face with near-infrared, not the
>>medium and far (thermal) infrared.  Near infrared is supposed to penetrate
>>flesh far better, so your blood vessels are visible and form a pattern
>>which can be recognized.
>>
>>Jim Bell
>>jimbell at pacifier.com
>
>Jim,
>
>Where did you get your info?  Near IR is around 1-1.5 microns, at these
>wavelengths, the body radiates very little energy.  I think most of the
>systems you are discussing use mid (3-5) or long-wave (8-12) IR, where
>objects that are room to body temp radiate most of their energy.

I get most of my information in this area from Photonics Spectra magazine, 
and Laser Focus World magazine.

No, Clay, I did not say that the flesh RADIATED near IR. (it does, but only 
a very tiny amount.)  The identification system I describe would probably 
use 940 nm IRLEDs to illuminate the face, and a silicon CCD detector to pick 
up the images.  Or it would use ambient near-IR, perhaps from the sun or a 
tungsten filament or fluorescent lighting, along with an IR filter to ensure 
that the CCD camera picked up only the IR bands of interest.  It would be 
easy to check out the results:  Put such an IR-passing filter in front of a 
CCD-based camcorder, and take a picture of somebody.

Incidentally, this simplicity shows the flaw in using this kind of system as 
an identifier:  Since people's faces are usually visible, and can be 
photographed in the near-IR surreptitiously, it isn't clear how to prevent 
faking a face which appears to have the same IR signature and pattern.

Jim Bell
jimbell at pacifier.com







More information about the cypherpunks-legacy mailing list