Thanks Daniel
I undestand... but... :)
the data that libs store of the image have some kind of representation... 
right?
or the lib allways compare image <-> image?

meaby if i can get some "data" about the finguerprint... like an array of 
curves descriptors and points... this match can be done faster?

i think i need to make a lot of research in order to find something.... 1st... 
i will read about how does the finguerprints are detected and what does the 
lib step by step to make a match...

can you (or someone) sendme some links to this kind of docs???

Thanks in advance!!!

On Tuesday 28 October 2008, Daniel Drake wrote:
> Jonatan Schijman wrote:
> >     I'm new using FinguerPrint Scan Technology.
> >     My goal... make a "very faster" system to autenticate mass of people
> > using finguers.... my idea... use a hash tree and make a huge memory
> > allocation using realtime linux options.
> >     It's posible to use libfprint to store a hash of the finguerprint and
> > then when we make the verification compare hashes only? but... with my
> > own hash store proc?
>
> How are you going to hash fingerprint images?
> No 2 fingerprint images are the same. Fingerprint authentication is the
> process of comparing two images for *similarity* (not equivalence), and
> thresholding that similarity at an appropriate level.
>
> This is in contrast to passwords/smartcards/etc where hashing is
> actually possible and makes a lot of sense.
>
> Daniel



-- 
El Rubio
_______________________________________________
fprint mailing list
[email protected]
http://lists.reactivated.net/mailman/listinfo/fprint

Reply via email to