[xcbf] [Fwd: Re: (U) RE: Standard field for Finger Index]

[Phil Griffin <phil.griffin@asn-1.com>]

FYI. Phil

-------- Original Message --------
Subject: Re: (U) RE: Standard field for Finger Index
Date: Fri, 19 Jul 2002 00:12:05 +0200
From: Alessandro Triglia <sandro@MCLINK.IT>
Reply-To: The Biometric Consortium's Discussion List 
  <BIOMETRICS@PEACH.EASE.LSOFT.COM>
To: BIOMETRICS@PEACH.EASE.LSOFT.COM

 > -----Original Message-----
 > From: The Biometric Consortium's Discussion List
 > [mailto:BIOMETRICS@PEACH.EASE.LSOFT.COM] On Behalf Of Phil Griffin
 > Sent: Thursday, July 18, 2002 23:15
 > To: BIOMETRICS@PEACH.EASE.LSOFT.COM
 > Subject: Re: (U) RE: Standard field for Finger Index
 >
 >
 > Cathy Tilton wrote:
 >
 > > In a message dated 7/18/2002 4:05:19 PM Eastern Daylight Time,
 > > Raymond.Purvis@LACKLAND.AF.MIL writes:
 > >
 > >
 > >>Does that mean it will de-facto roll into BioAPI 1.X and
 > X9.84?  Thanks.
 > >>
 > >
 > > No.  There is no plan at the current time to include it in
 > BioAPI.  It was
 > > considered early on in the development and considered
 > unnecessary (to the
 > > application, the generating/receiving technology could
 > include it in the
 > > opaque data subheader, if needed).
 > >
 > > There is a revision of X9.84 in progress, so it could be
 > included in there as
 > > an optional header field, but if the data were originally
 > generated by a
 > > BioAPI BSP, it could not be populated.
 >
 >
 > There's the rub it seems. You can map the information
 > in a BioAPI BIR record to an X9.84 BiometricObject, but
 > you can't map everything in a BiometricObject into a BIR.
 >
 > Ideally, one would define a set of BioAPI call backs that
 > used a BiometricObject in place of a BIR. Or alternatively,
 > X9.84 would define a set of BioAPI-like call backs.
 >
 > Either way you would end up with a complete end to end
 > technical solution that was arguably better than either
 > the current BioAPI or X9.84 standards provide. A common
 > record format whose biometric information could be secured
 > using the cryptographic processing, key management and
 > security requirements defined in X9.84.


I agree.  Today, when one thinks of a biometrics system architecture
including both BioAPI and X9.84, the latter is usually viewed as
addressing security issues for the *transmission* of biometric 
samples
and templates.  Unfortunately, once the data is received or 
before it is
sent, it lives within the domain of BioAPI and outside the 
applicability
of X9.84.  However, for many applications it would be beneficial if
these two standards were integrated more tightly.

One solution would be to extend the BioAPI standard to allow the 
use of
an X9.84 data block (an "encoded" BiometricSyntax value) in place 
of a
BIR (or as a new kind of BIR), as Phil suggests.  Among the 
benefits is
the possibility of storing in a template database a digitally-signed
biometric template that keeps the signature of the original 
signer, with
which the data was received.  The signature could then be checked 
at any
time, both when the template is received from the network and 
when it is
retrieved from the database months or years later.

Alessandro Triglia
OSS Nokalva


-------------------------------------------------------------------
The preceding was forwarded by the Biometric Consortium's Electronic
Discussion Group.  Any opinions expressed here do not necessarily
reflect those of the Biometric Consortium.  Further distribution
is prohibited.

Problems and questions regarding this list should be sent to
BIOMETRICS-request@PEACH.EASE.LSOFT.COM.

To remove your name from this list please send the command
"SIGNOFF BIOMETRICS" to <LISTSERV@PEACH.EASE.LSOFT.COM>.  Please
do not send the "SIGNOFF BIOMETRICS" command to the BIOMETRICS list.

To update membership information (new e-mail address etc.), 
please send
a message to <bailey@biometrics.org> providing the updated 
information.
-------------------------------------------------------------------