ATVS logo Biometric Recognition Group - ATVS EPS logo UAM logo

INSTRUCTIONS FOR DOWNLOADING ATVS-FFp DB

  1. Download license agreement, send by email one signed and scanned copy to atvsuam.es.
     

  2. Send an email to atvsuam.es, as follows:
    Subject: [DATABASE download: ATVS-FFp DB]

    Body
    : Your name, e-mail, telephone number, organization, postal mail, purpose for which you will use the database, time and date at which you sent the email with the signed license agreement.
     

  3. Once the email copy of the license agreement has been received at ATVS, you will get an email with a username, a password, and a time slot to download the database.
     

  4. Download the database, for which you will need to provide the authentication information given in step 4. After you finish the download, please notify by email to atvsuam.es that you have successfully completed the transaction.
     

  5. For more information, please contact: atvsuam.es



DESCRIPTION OF ATVS-FFp DB

Two datasets of real and fake fingerprint images in .bmp format:


    DATASET 1: DS_WithCooperation

The gummy fingers from which the fake fingerprint images were taken, were generated with the cooperation of the user according to the methodology described in [TS2010].

It contains fingerprint samples of the index and and middle fingers of both hands of 17 users (17 x 4 = 68 different fingers).

Four samples of each fingerprint (fake and real) were captured in one acquisition session with:

  1. The flat optical sensor Biometrika Fx2000 (512 dpi)

  2. The sweeping thermal sensor by Yubee with Atmel's Fingerchip (500 dpi)

  3. The flat capacitive sensor by Precise Biometrics model Precise 100 SC (500 dpi).

This way the dataset comprises 68 fingers x 4 samples x 3 sensors = 816 real image samples and as many fake images.


    DATASET 2: DS_WithoutCooperation

The gummy fingers from which the fake fingerprint images were taken, were generated with the cooperation of the user according to the methodology described in [TS2010].

It contains fingerprint samples of the index and and middle fingers of both hands of 17 users (16 x 4 = 64 different fingers).

Beware that user 14 is missing.

Four samples of each fingerprint (fake and real) were captured in one acquisition session with:

  1. The flat optical sensor Biometrika Fx2000 (512 dpi)

  2. The sweeping thermal sensor by Yubee with Atmel's Fingerchip (500 dpi)

  3. The flat capacitive sensor by Precise Biometrics model Precise 100 SC (500 dpi).

This way the dataset comprises 64 fingers x 4 samples x 3 sensors = 768 real image samples and as many fake images.


    NOMENCLATURE

The nomenclautre followed in both datasets to name the image files is as follows: uXX_A_BB_CD_YY

  • XX: is the number of the user [01 02 03 ... 17]

  • A: it can take the values "o" or "f" for "original" or "fake"

  • BB: it can take the vaules "fc", "fo", or "ft", for "capacitive", "optical", or "thermal" depending ono the sensor used to capture the image

  • C: it can take the values "r" or "l", for "right" or "left" hand

  • D: it can take the values "m" or "i", for "middle" or "index" finger

  • YY: is the number of the sample [01 02 03 04]


    CORRESPONDENCE BETWEEN DATASETS

Users number [03 04 05 09 13 15 16 17] are the same in both datasets.



REFERENCES

For further information on the database and on different applications where it has been used, including vulnerability assessment of fingerprint verification systems to direct attacks, and performance evaluation of fingerprint liveness detection methods, we refer the reader to:

Please remember to reference article [FGCS2012] on any work made public, whatever the form, based directly or indirectly on any part of the ATVS-FFp DB.

 
Sample fingerprint 1 - Real capacitive sensor

Sample fingerprint 2 - Real optical sensor

Sample fingerprint 3 - Real thermal sensor

Sample fingerprint 4 - Fake capacitive sensor

Sample fingerprint 5 - Fake optical sensor

Sample fingerprint 6 - Fake thermal sensor