Unfortunately, that will not work. Correct me if I'm wrong, but a
fingerprint is never (for all intents and purposes) the same scanned twice
in a row. The magic matching of scanned fingerprint against stored prints is
performed by the Bozorth3 algorithm based on discovered minutiae. And the
match decision is not a binary "matched"/"not matched" answer, but rather a
matching score.

Hence you will never be able to reproduce the encryption key (or message for
the digest), which renders any systems based on this approach quite useless.

-Artem

On Mon, Jun 1, 2009 at 8:51 PM, <[email protected]> wrote:

> Hi..
>
> At the end of
> http://reactivated.net/fprint/wiki/Security_notes#Disk_storage ,
> the problem of encrypting fingerprints on disk was raised.
>
> I've got a solution: use the fingerprint as a key to encrypt a fixed
> string.
>
> This is what the unix password system used for ages.
>
> Alternatively, hash the fingerprint with md5, sha1, or whatever you want.
>  This
> is what the current unix password system does, using PAM.
>
> If the hash of a new fingerprint matches the hash of the enrolled
> fingerprint,
> they're the same fingerprint (to a very high probability).
>
> For even higher security, pick some random letters to prepend to the
> fingerprint data, hash it, and store the hash and the random letters.  It's
> designed to prevent two databases from being compared to see if the same
> fingerprint is in both.
>
> Good luck!
> _______________________________________________
> fprint mailing list
> [email protected]
> http://lists.reactivated.net/mailman/listinfo/fprint
>
_______________________________________________
fprint mailing list
[email protected]
http://lists.reactivated.net/mailman/listinfo/fprint

Reply via email to