headers
Robert Vanak | 1 Dec 2009 16:56
Picon

New matcher project

Hi all,

I am starting new opensource project for fingerprint recognition and
matching. Before starting however, I would like to request your
feedback, especially:

* Are there any other opensource matchers besides Bozorth3 and FVS?
* Are there any public fingerprint databases besides FVC sample DBs,
Neurotech DBs, and the artificial SFinGe output?
* Do you see anything silly about the project? Something that I overlooked?

I intend the project to be in many ways better than Bozorth3 and FVS,
specifically:

* Faster, 1,000s or 10,000s of matched fingerprints/second.
* More accurate through use of ridge matching besides minutia
matching. Checked in FVC-onGoing.
* Customizable with parameters, accuracy/speed tradeoffs, DB-driven
optimizer, and clean editable sources.
* More stable and easier to develop by prototyping in C# and porting
to C++ and Java.

The project sources and other data will be published here:
https://sourceforge.net/projects/sourceafis/

Thanks for your feedback,
Robert
Permalink | Reply | 4 comments |
headers
Daniel Drake | 1 Dec 2009 19:17
Favicon

Re: New matcher project

Robert Vanak wrote:
> Hi all,
> 
> I am starting new opensource project for fingerprint recognition and
> matching. Before starting however, I would like to request your
> feedback, especially:

Great news!

> * Are there any other opensource matchers besides Bozorth3 and FVS?

Not that I know of.

> * Are there any public fingerprint databases besides FVC sample DBs,
> Neurotech DBs, and the artificial SFinGe output?

Is anything here of use?
http://fingerprint.nist.gov/NBIS/index.html

> * Do you see anything silly about the project? Something that I overlooked?

It sounds excellent. In fact, bozorth3 is one of the biggest sticking 
points for the fprint project at this time. We'd all be very happy if 
you could improve on this.

The problems are that it works very well for fingerprints of young 
people, but not so well for the older generations where the skin is looser.

The other big problem is that it doesn't work well for images with very 
few minutae (less than 10), and some of the sensors that are common in
(Continue reading)

Permalink | Reply | 3 comments |
headers
Robert Vanak | 1 Dec 2009 20:04
Picon

Re: New matcher project

On Tue, Dec 1, 2009 at 7:17 PM, Daniel Drake <dan <at> reactivated.net> wrote:
Robert Vanak wrote:
* Are there any other opensource matchers besides Bozorth3 and FVS?

Not that I know of.

Further research turned out "Fingerprint recognition SDK" which seems much better than Bozorth3, judging from its documentation. I am reconsidering my plans in view of tougher competition. It might be a better idea to improve on top of existing FR-SDK sources, but then that's not so much fun. I have compiled a list of other algorithms with quick evaluation of each here:

https://sourceforge.net/apps/mediawiki/sourceafis/index.php?title=Main_Page

Is anything here of use?
http://fingerprint.nist.gov/NBIS/index.html

There appears to be an image database, but I have no idea in what format are those AN2 files.
 
The problems are that it works very well for fingerprints of young people, but not so well for the older generations where the skin is looser.
 
The other big problem is that it doesn't work well for images with very few minutae (less than 10), and some of the sensors that are common in the marketplace only produce small images like this.

Both these problems plague all matching algorithms. Both can be solved to a large extent with better scanning devices, but I understand this is not an option all the time.

Robert

_______________________________________________
fprint mailing list
fprint <at> reactivated.net
http://lists.reactivated.net/mailman/listinfo/fprint
Permalink | Reply | 2 comments |
headers
Daniel Drake | 1 Dec 2009 22:21
Favicon

Re: New matcher project

Robert Vanak wrote:
> Further research turned out "Fingerprint recognition SDK" which seems 
> much better than Bozorth3, judging from its documentation. I am 
> reconsidering my plans in view of tougher competition. It might be a 
> better idea to improve on top of existing FR-SDK sources, but then 
> that's not so much fun. I have compiled a list of other algorithms with 
> quick evaluation of each here:

Never heard of that one before.

> There appears to be an image database, but I have no idea in what format 
> are those AN2 files.

The PDF documentation is pretty good.

> Both these problems plague all matching algorithms. Both can be solved 
> to a large extent with better scanning devices, but I understand this is 
> not an option all the time.

Either way, the proprietary solutions that come with the hardware device 
themselves usually work OK, whereas fprint often fails for certain 
fingers/setups.

Daniel
Permalink | Reply | 1 comment |
headers
Sarath Lakshman | 5 Dec 2009 13:12
Picon

Fingerprint based file protection

Hi,

I am trying to implement fingerprint based fileprotection.
I need to generate a unique key for a fingerprint scan. So that I can
use that key for encryption.
In the next scan, using the key generated we can decrypt the file.

Is it possible to implement file protection using libfprint ?

Happy Hacking,
Sarath
Permalink | Reply | 3 comments |
headers
Pavel Herrmann | 5 Dec 2009 14:27
Picon

Re: Fingerprint based file protection

Hi

this is not possible, or at least not in an easy way
fingerprints are checked for similarity, not equality, so you would need some 
kind of algorithm to make the same key for similar fingerprints - which is not 
possible, if you have two different fingerprint there could be one that is 
similar to both (is half-way between them), thus you would have the same keys 
for different fingerprints

Pavel

On Saturday 05 of December 2009 13:12:44 Sarath Lakshman wrote:
> Hi,
> 
> I am trying to implement fingerprint based fileprotection.
> I need to generate a unique key for a fingerprint scan. So that I can
> use that key for encryption.
> In the next scan, using the key generated we can decrypt the file.
> 
> Is it possible to implement file protection using libfprint ?
> 
> 
> Happy Hacking,
> Sarath
> _______________________________________________
> fprint mailing list
> fprint <at> reactivated.net
> http://lists.reactivated.net/mailman/listinfo/fprint
>
Permalink | Reply | 0 comments |
headers
Robert Vanak | 5 Dec 2009 17:00
Picon

Re: New matcher project

On Tue, Dec 1, 2009 at 10:21 PM, Daniel Drake <dan <at> reactivated.net> wrote:
Robert Vanak wrote:
Further research turned out "Fingerprint recognition SDK" which seems much better than Bozorth3, judging from its documentation. I am reconsidering my plans in view of tougher competition. It might be a better idea to improve on top of existing FR-SDK sources, but then that's not so much fun. I have compiled a list of other algorithms with quick evaluation of each here:

Never heard of that one before.

Okay, I investigated Fingerprint recognition SDK in more detail and here's what I have found:

* much newer than Bozorth3, it's just 2 years old
* sound algorithms that are a considerable improvement over Bozorth3
* more accurate than Bozorth3, uses non-minutia information
* faster than Bozorth3, should be practical to search databases of 10,000 fingerprints
* tunable if provided with database of training fingerprints, just give it fingerprints from fprint-supported sensors
* tuning can be also used to specialize the algorithm for specific sensors, especially the small ones
* high tolerance to elastic deformation of fingerprints, this might be useful for the elderly

I haven't tested it though, but I guess the tests conducted by authors can be trusted.

See more information about their algorithm here:
http://fingerprintreco.cvs.sourceforge.net/viewvc/fingerprintreco/fingerprintreco/Documents/Presentations/

So my recommendation is to integrate Fingerprint recognition SDK in fprint as a replacement for Bozorth3.

Regarding my own project, I do not expect it to provide any revolutionary improvement over FR-SDK. I might do it for the fun anyway, I haven't yet decided about this.

Robert

_______________________________________________
fprint mailing list
fprint <at> reactivated.net
http://lists.reactivated.net/mailman/listinfo/fprint
Permalink | Reply | 0 comments |
headers
Gustavo Chaín Dumit | 7 Dec 2009 13:02
Picon

Re: Fingerprint based file protection

El Sat, 5 Dec 2009 17:42:44 +0530
Sarath Lakshman <sarathlakshman <at> gmail.com> escribió:
> Hi,
Hi !

> 
> I am trying to implement fingerprint based fileprotection.
> I need to generate a unique key for a fingerprint scan. So that I can
> use that key for encryption.
You can use a database to store [fingerprint|unique key]
That way, when finger match against fingerprint stored on DB, you will
get the unique key.
> In the next scan, using the key generated we can decrypt the file.
> 
> Is it possible to implement file protection using libfprint ?
> 
> 
> Happy Hacking,
> Sarath
> _______________________________________________
> fprint mailing list
> fprint <at> reactivated.net
> http://lists.reactivated.net/mailman/listinfo/fprint

--

-- 
Gustavo Chaín Dumit
http://0xff.cl
_______________________________________________
fprint mailing list
fprint <at> reactivated.net
http://lists.reactivated.net/mailman/listinfo/fprint
Permalink | Reply | 1 comment |
headers
Pavel Herrmann | 7 Dec 2009 13:33
Picon

Re: Fingerprint based file protection

Hi,

On Monday 07 December 2009 13:02:29 Gustavo Chaín Dumit wrote:
> El Sat, 5 Dec 2009 17:42:44 +0530
> 
> Sarath Lakshman <sarathlakshman <at> gmail.com> escribió:
> > Hi,
> 
> Hi !
> 
> > I am trying to implement fingerprint based fileprotection.
> > I need to generate a unique key for a fingerprint scan. So that I can
> > use that key for encryption.
> 
> You can use a database to store [fingerprint|unique key]
> That way, when finger match against fingerprint stored on DB, you will
> get the unique key.
> 

this way the database would have to be encrypted and the matching application 
would have to be "trusted" in the way that it knows the key to decode the 
database (without user intervention, that means hardcoded key) and then do the 
matching/retrieving the file decryption key

as long as you are OK with hardcoding decryption keys into the app this is a 
correct solution, but it is not as safe as one could think

Pavel
_______________________________________________
fprint mailing list
fprint <at> reactivated.net
http://lists.reactivated.net/mailman/listinfo/fprint
Permalink | Reply | 0 comments |
headers
Aaron W. Hsu | 9 Dec 2009 23:00
Picon
Gravatar

AuthenTec readers?

Hello everyone,

I was just wondering if any progress has been made recently on the
AuthenTec series of finger print readers, like those found on the Lenovo
T500 (one of which I happen to own). It would be great if these worked,
but I don't see any Wiki updates except for a first-analysis mentioned.

  Aaron W. Hsu
Permalink | Reply | 0 comments |

Gmane