SlideShare a Scribd company logo
1 of 7
Download to read offline
Image Quality and Minutiae Count Comparison
             for Genuine and Artificial Fingerprints
            Stephen J. Elliott Ph.D., Shimon K. Modi, Lou Maccarone, Matthew R. Young, Changlong Jin, and
                                                     Hale Kim Ph.D.

                                                                                  same user after acquisition on a commercially available
   Abstract-The vulnerabilities of biometric sensors have been                    biometric fingerprint sensor. This research seeks to answer
discussed extensively in the literature and popularized in films and              whether the samples (live versus gelatin) exhibit the same
television shows. This research examines the image quality of an                  minutiae counts and whether the acquired images possess the
artificial print as compared to a genuine finger, and examines the                same image quality properties. That is, are the live and gelatin
characteristics of the two, including minutiae counts and image
quality, as repeated samples are taken.                                           samples similar in their characteristics, and do their
                                                                                  similarities or differences have an impact on matching
   Keywords-authentication, biometrics, fingerprint recognition,                  perfonnance?
repeatability.
                                                                                               II. BIOMETRIC SYSTEM VULNERABILITIES
          I. INTRODUCTION AND MOTIVATION                                             All security measures, including mechanisms for
  V ERIFYING the identity of an individual can be                                 authenticating identity, can be circumvented. The processes
    Vaccomplished by three main considerations: what an                           associated with working around these measures vary in
individual has, what an individual knows or owns, and what                        difficulty according to the level of effort and resources needed
an individual is. The first option is typically achieved through                  to cany out the deceptive act. Authentication mechanisms
the use of a token, such as an identification card, badge,                        based on secrets are particularly vulnerable to "guessing"
magnetic stripe, or radio frequency identification (RFID) tag.                    attacks. Token mechanisms that rely on the possession of an
The second option can be achieved through the use of a                            object, typically a card or badge technology are most
password or personal identification number (PIN). The third                       vulnerable to theft or falsified reproduction. Biometric
option can be accomplished through what an individual is by                       technologies closely tie the authenticator to the individual
utilizing biometric technologies. Like the first two                              identity of the user through the use of physiological or
authentication methods, biometric systems contain                                 behavioral characteristics. While this property offers an added
vulnerabilities and are susceptible to attack. Some of these                      advantage over the other two authentication mechanisms, it
vulnerabilities are similar or even overlapping across all three                  places a great emphasis on validating the integrity of the
authentication mechanisms. However, attacks specific to                           biometric sample acquired and transferred in the biometric
biometric systems focus on liveness detection of a human (i.e.,                   system. Ratha, Connell, and Bolle provided a model
whether the finger from a live sample or a gelatin sample).                       identifying vulnerabilities in biometric systems [3]. An
Various attacks documented in the literature have focused on                      example of the threat model is shown in Fig. 1, and builds on
the sensor [11, [2]. While understanding and preventing                           the general biometric model outlined in Mansfield and
attacks on the sensor is an interesting research topic worthy of                  Wayman [4].
investigation, this paper examines the global and local features
of a live sample compared to that of a gelatin finger from the

   S. J. Elliott, Ph.D. is an Associate Professor with the Departnent of
Industrial Technology at Purdue University, West Lafayette, IN 47907 USA;
(e-mail:elliott apurdue.edu .
   S. Modi is graduate student with the Departnent of Industrial Technology
at Purdue University, West Lafayette, IN 47907 USA; (e-mail:shimon@
purdue.edu).
   L. Maccarone is an undergraduate student at Purdue University, West
Lafayette, IN 47907 USA                                                                             Fig. 1 Biometric system threat model
   M. R. Young is a graduate student with the Department of Industrial
Technology at Purdue University, West Lafayette, IN 47907 USA
   C. Jin is a graduate student with the School of Information and             The biometric system threat model shown in Fig. 1 contains
Communication Engineering at Inha University, Incheon, Korea; (e-            11 individual areas of vulnerability. In addition to the five
mail:cljin@vision.inha.ac.kr).                                              main internal modules characterized in the general biometric
   H. Kim is a Professor with the School of Infornation and Communication
Engineering at Inha University, Incheon, Korea; (e-mail:hikim@inha.ac.kr).  model (data collection, signal processing, matching, storage,
    1-4244-1129-7/07/$25.00 ©2007 IEEE                                     30

            Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
and decision) [4], another component is added to represent the                   the features of the two fingerprints remain consistent
transfer of the authentication decision to the application that                  Repeatability of the extracted features is important for the
relies on the decision from the, biometric system. Such                          matching process in any type of biometric technology [7]. The
applications could be identity management systems (IDMS) or                      features to be examined include: minutiae points and image
access control systems for logical and or physical access to                     quality. One of the challenges associated with this research is
resources. These systems can vary in complexity and size,                        to ensure the image is of sufficient quality. A wide variety of
ranging from a local computer log-in all the way to a wide-                      factors can influence the quality of fingerprint samples. Non-
scale distributed architecture seen in the cases of the U.S.                     uniform contact, inconsistent contact, or irreproducible contact
Department of Transportation's Transportation Worker                             with the fingerprint sensor can result in images with a low
Identification Credential (TWIC) [5] or the Personal Identity                    signal-to-noise ratio, which is not desirable for feature
Verification (PIV) of Federal Employees and Contractors [6].                     extraction and matching purposes [9]. Wear and tear of the
The remaining points of vulnerability are communication                          skin and aging effects can semi-permanently alter ridge
channels between these six modules. It is worth noting that not                  characteristics. These factors also affect acquisition of
all 11 vulnerability points are unique to the biometric system.                  fingerprints by the fingerprint sensor. The importance of
Many of the same points (e.g., storage and communication                         quality is widely acknowledged, but there is no standard
channels) are vulnerable in other authentication systems and                     means of assessing quality. The current standardization effort
similar methods can be used to limit those particular                            for assessing quality for biometric samples refers to three
vulnerabilities.                                                                 different connotations of quality:
   The most publicized vulnerability in biometric systems                             * Character
resides at the data collection module in the form of spoofmg or                       * Fidelity
presenting artificial representations of biometric samples (see                       * Utility
module #1, Fig. 1). If an artificial or fake biometric sample is                    These three connotations of biometric sample quality can be
accepted by the biometric system at this initial stage, the entire               directly applied to fingerprint sample quality. Character is a
biometric system is corrupted and the system is compromised.                     description of quality based on inherent features from the
Attacks on biometric systems are not new, popular culture                        source of the fingerprint. Individuals who have scarred
seeks to circumvent security systems and biometric systems                       fingerprints or dry or cracked skin on the fingertips will
are not immune. Several online resources are available that                      provide samples with poor character. Fidelity is a description
describe such attacks on the data collection module, and many                    of quality based on degree of similarity between the actual
movies and television shows have highlighted attacks on such                     fingerprint and the fingerprint image acquired by the sensor.
systems. One such attack at this data collection module was                      Inconsistent contact with the fmgerprint sensor can lead to
outlined in the work of Matsumoto, Yamada, and Hoshino                           fingerprint samples with poor fidelity. Utility is a description
(2002) using "Gummy Fingers" [1].                                                of quality based on observed or predicted contribution of the
   The biometric research community, as well as industry, has                    fingerprint sample to the overall performance of the
focused its research on preventing such attacks by using the                     fingerprint recognition system. Utility of a fmgerprint sample
concept of "liveness" detection techniques. Today, the newer                     is directly affected by the character and fidelity of the
sensors are improving their resilience against a spoofing attack                 fingerprint sample, and should be the closely related to
at this module. In the past, acetate spoofing attacks - where                    performance of the recognition system.
an image of a fingerprint placed on acetate was accepted as a                       A substantial amount ofresearch has been conducted in area
genuine live finger - was easy to reproduce. Now, such                           of quality assessment, all of which give varying levels of
attacks are proving increasingly difficult to succeed, hence the                 importance to character, fidelity, and utility. Previous research
more complicated approaches to attacks being waged on the                        in the field of fingerprint image quality assessment can be
vulnerable sensor. Techniques for liveness detection within                      generalized into three categories: local features analysis,
the fingerprint modality focus on moisture content,                              global features analysis, and quality analysis as a classification
temperature, electrical conductivity, and challenge response.                    problem [10]. Features of the fingerprint image such as
                                                                                 minutiae count, fidelity of minutiae, contrast ratio between
         III. FINGERPRnT IMAGE QUALITY ANALYSIS                                  ridges and valleys, capture area of the figerprint, and
   The purpose of this research paper was not to prove the                       detennination of dominant direction are used by quality
vulnerability of the biometric system, but to examine the                        algorithms in varying capacities to make quality
repeatability of the features of the gelatin finger print as                     determinations. For the purpose of this research, fingerprints
compared to the live genuine sample once the image has been                      from live fngers and gelatin fingers were examined by two
acquired. The research question is whether an artificial print                   different image quality algorithms, one provided by Aware,
captured on an optical sensor exhibits any of the same                           Inc. and the other provided as a part of NIST Fingerprint
characteristics as a genuine fingerprint from the same                           Imaging Software (NFIS).
individual captured on the same sensor, and whether any
distinguishers might enable the artificial print to be excluded                                            IV. METHODOLOGY
later on in the process, if the initial data collection module                      This research involved two separate experiments. The first
accepts it. The research also investigates whether, over time,
                                                                               31

           Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
experiment was conducted in two stages. Stage 1 of the first                    verified by the software, then the mold would have been
experiment involved creation of a set of images from an                         destroyed and the process started again.
artificial gelatin fimger that was crafted from the same subject                   The artificial finger was returned to the refrigerator in a
who provided the genuine fingerprint. The procedures to                         simple (airtight, but not vacuum-sealed) plastic storage
accomplish this feat required adaptation of several different                   container for 48 hours at a temperature of 2TC. This procedure
methodologies outlined in the literature for creating an                        allowed the gelatin finger to completely solidify to its
artificial fingerprints, including the work done by Matsumoto,                  permanent state.
Yamada, and Hoshino [1]. Prior to creating the mold, the                           After removing the gelatin finger from the refrigerator, tests
following necessary ingredients and utensils were gathered:                     were conducted on it to estimate the optimal load required for
plastic clay, hot water, and a pair of plastic tongs. To create                 acquiring images. In general terms, it is best to use the least
the mold, a quantity of plastic clay sufficient to cover the                    weight possible to produce a scan in order to minimize the
genuine finger was required. In order to make the plastic clay                  spreading and dissolution of the gelatin finger's valleys and
malleable, it was placed briefly in boiling water. In order to be               ridges. Testing of loads ranging from 20Og to 1,000g, as
utilized, the plastic clay had to have a consistency such that it               measured by a Tanita digital scale, was performed.
enabled a mold to be created by placing the finger with only                    Approximately 200g was determined to be the lower limit to
light pressure. When the plastic clay has attained a sufficient                 produce an image, with 550g being the upper limit before
level of pliability, the clay was cooled. Once the clay had                     distortion and inability to match occurred. The next stage of
cooled enough to be touched, the finger was placed into the                     the experiment involved acquisition of a series of images from
                                                                                the gelatin finger. All of the images were acquired from the
clay to a depth sufficient to create the mold to be used to craft               optical sensor using the gelatin fimger over a 15-minute time
the gelatin finger. The genuine finger was kept in the plastic                  period. After 15 minutes of acquiring images, the gelatin
clay until the clay had cooled enough to retain its shape and                   finger had degraded to the point at which it was no longer able
the details of the genuine finger. After the finger was                         to be accepted by the optical sensor. In all, 163 images were
removed, the mold was allowed to cure for an additional 10                      produced over the 15-minute time period. A detailed
minutes. The resulting mold for this study is shown in Fig. 2.                  description of the 160 images utilized in the study is provided
                                                                                in the results section.
                                                                                   Stage 2 of the first experiment called for the collection of a
                                                                                series of live samples. One hundred sixty live samples were
                                                                                acquired from the same finger that was used to create the
                                                                                artificial gelatin finger. The live samples were collected over
                                                                                an 8-minute time period on a commercially available optical
                                                                                sensor. The authors chose to collect 160 live samples, as this
                                                                                was the same number of fingerprints collected from the gelatin
                                                                                finger. These 160 live images were all stored according to the
                                                                                time collected; they are used to provide a baseline quality
                                                                                assessment that will be compared against the samples
         Fig. 2 Plastic mold formed to create gelatin finger                    generated by the gelatin finger.
                                                                                   After data collection, both sets of images (from the live
   The next step was to create a gelatin mixture capable of                     finger and the gelatin fimger) were processed through the NIST
producing artificial fingerprints from the mold that would be                   Fingerprint Image Software (NFIS) package. The MINDTCT
recognizable to the sensor. Two sheets of gelatin weighing                      function was used to count the number of minutiae present in
3.5g were soaked in cold water for five minutes. In order to                    each individual image. The NFIQ function was used to
remove the excess water, the gelatin sheets were dried until                    evaluate image quality, which is determined on a rating scale
the gelatin weighed 14-16g. Next, a bowl was immersed into                      of 1 to 5, with 1 being the best and 5 being the worst. The
extremely hot water, and the gelatin was placed into the bowl                   results of M1NDTCT and NFIQ from both groups (the live
to soften and melt the gelatin. Once the gelatin had melted, it                 fingerprints and the gelatin fingerprints) were then compared
was poured into the clay mold. Immediately after placing the                    by the means of statistical t-tests (using an a level of 0.05) to
gelatin in the mold, the mold was placed in a refrigerator to                   determine if any statistical difference existed across the
cool for 10 minutes at a temperature of 1°C. Cooling                            groups. Aware, Inc. offers a commercially available image
transforms the gelatin to a state that is resistant to changes in               quality and minutiae count software; this software was used to
shape when touched. Ambient room temperate, when this                           extract image quality scores and minutiae counts for
experiment was conducted, was 220C.                                             fingerprints from the live finger and gelatin finger groups. By
   After the gelatin finger had cured in the refrigerator for a                 utilizing two different software packages to analyze the
period of approximately one hour, it was removed from the                       fingerprints, we sought to eliminate bias that might have been
clay mold and placed on the sensor to determine whether it                      generated by utilizing only a single application.
was actually able to produce images. Ten attempts to acquire                       The second experiment involved collecting fingerprint
an image were made; in all 10 instances, an image was                           images from left index finger and the right thumb from 30
produced. The software verified the mold. If the mold was not                   different subjects. Each subject was asked to provide 20
                                                                              32

          Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
images of each fimger on three different optical sensors. A                      of live and gelatin fingerprint minutiae count for last 16 prints.
silicon mold was created the left index and right thumb of                       An interesting observation is that the minutiae count increases
each subject, and the molds were used to provide 20                              for the gelatin fingerprint group, but stabilizes for live
fingerprint images on each of the three optical sensors. The                     fingerprints.
procedures used in [1] were also employed in creating the
silicon molds in the second part of the experiment. At the end                                      Boxplot of AWAREJLive-comt, AWARE_Gelatincount
of the data collection there were 3600 fingerprint images from                           70.
live fingers, and the 3600 fingerprint images from silicon
fingers.
                                                                                         60*


  A. Experiment I
                             V. RESULTS
                                                                                     a
                                                                                         50I
Creation and acquisition of images from gelatin fingers can be
problematic, as previous research has shown that gelatin                                 40     _
fingers do not afford consistent repeatability. However, this
study provides anecdotal evidence suggesting that better                                 3J
preparation and storage of the artificial finger can aid in the
repeatability of the images produced. The first 39 samples                                              AWARE Live ca,t            AWARELGelItkSem_t
provided consistently successful spoofing results; on the 40th                      Fig. 4 Box plot, live and gelatin finger minutiae count using Aware
presentation of the gelatin finger, a failure-to-acquire (FTA)                                                  QualityCheck
resulted. Overall, 163 images were acquired, but only 160
images were used for the final study. Degradation on the final
3 images rendered the images unusable. The acquisition rate                                          Boxplot of t.S_Live_Coumt, NFI&Gelatin cout
for this particular gelatin fingerprint was 90.7%, producing a                           110


FTA rate of 9.3%. Fig. 3 shows the gelatin print (left) a live
print (right). The FTA rate for the live finger was 0.0%.
                                                                                         90-




                                                                                          900
                                                                                    a
                                                                                         0-


                                                                                         60-
                                                                                    o
                                                                                         60

                                                                                         50
                                                                                         40

                                                                                                          NRSLve_Cort               NFS.Geeatincount

                                                                                    Fig. 5 Box plot, live and gelatin finger minutiae count using NFIS
                                                                                                                M1NDTCT

          Fig. 3 Gelatin finger (left) and live fmger (right)                       The stabilization of minutiae count for the live fingerprints
                                                                                 can probably be attributed to habituation or acclimation to the
   The minutiae count analysis on both the fingerprint groups                    device. The subject has been acclimated to placing the sample
was   performed. Fig. 4 and Fig. 6 show box plots of the                         finger on the optical sensor, which reduces the inconsistent
minutiae count from the live finger and gelatin finger,                          contact of finger surface with the platen of the sensor. Another
respectively, generated from Aware, Inc.'s image quality tool.                   interesting observation is the increase in minutiae count
Fig. 5 and Fig. 7 show box plots of the minutiae count from                      between the gelatin fingerprints over time. This suggests
the live finger and gelatin fmger, respectively, generated from                  degradation of the gelatin finger and mold because of
NFIS's MINDTCT.                                                                  repetitive use and introduction of cracks in the mold used to
   The results from the box plot graphs generated by both the                    create the gelatin finger. Evidence suggests that, over time, the
Aware and NFIS software programs show that the live                              number of minutiae for the gelatin fingerprint increases, while
fingerprints have a lower minutiae count than the gelatin                        the number of minutiae for the live fmgerprint stabilizes.
fingerprints, which is most likely a result of indirect and
inconsistent contact with the optical sensor. In order to study
the deterioration of the gelatin fingerprints, the first 16 and last
16 samples from the live and gelatin fmgerprint groups were
used. Fig 6 shows a box plot of the live and gelatin fingerprint
minutiae count for first 16 prints, and Fig. 7 shows a box plot
                                                                               33


           Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
9
      55

      501

      45

      40


      35

      303




      65-


      So
      60-

      55-



      45-

      40-

      35-

      30
                       Boxplot of Lisfe_First-16;, Gelatin-First-16




                                   I

                           ve-FirsL6




                       Boxplot of Live_Last_6, GelatinjLastL16




                          Live_Last_16
                                                'I'--



  Fig. 8 and Fig. 9 shows the scatter plots for minutiae count
versus sample numbers of live and gelatin fingerprint groups.
Both of these graphs give credence to the observations made
from the box plots.

  Scatterplot of NFISJJve-Count, WISjGelatin_count vs rnageCount




  >
      110

      100



       80

       70
      6.0

       40
            0     2
                   0
                    I
                      1



                  S%mN&



                          40
                        me_pmft
                               U



                               .



                               Jj,#MO

                                   6
                                       U




                                       *-
                                            0
                                                   "



                                                100 12
                                                         1




                                                         14
                                                              U




                                                              1


 Fig. 8 Scatter plot, minutiae count vs. sample number using NFIS
                                                MIINDTCT
                                                                  Gebtin_Fst16
 Fig. 6 Box plot, live and gelatin finger minutiae count using Aware,
                             first 16 prints




                                                                  Gelati_Last_16

 Fig. 7 Box plot, live and gelatin finger minutiae count using Aware
                     QualityCheck, last 16 prints




                                                                         -
                                                                                VariabLa
                                                                        -+- EIS Lwe.Co.wA
                                                                                           -'-




                                                                             W- NFIS_GeaUn_cwnt
                                                                                                        (U
                                                                                                        'U
                                                                                                        9




                                                                                                   Groups
                                                                                                           70

                                                                                                             60-


                                                                                                             50-


                                                                                                             40-


                                                                                                             30-




                                                                                                             70
                                                                                                             so 80



                                                                                                             40




                                                                                                   Aware_LiveLIQ
                                                                                                                   0




                                                                                                   Aware_Gelatin_IQ 160
                                                                                                   NFIS_Live-IQ     160
                                                                                                   NFIS Gelatin IQ 160
                                                                                                                            *




                                                                                                                            20
                                                                                                                                 tE a




                                                                                                                                     ...
                                                                                                       Scatterplot of AWAREJive_ount, AWARE_Gelatincmnt vs bnaJeCoLnt
                                                                                                                                 ,V,able




                                                                                                                                     ,

                                                                                                                                     40




                                                                                                                                     5




                                                                                                                                    U0
                                                                                                                                 0406
                                                                                                                                          a




                                                                                                                                              60




                                                                                                                                              *
                                                                                                                                                  ILI




                                                                                                                                                  *rn
                                                                                                                                                    *




                                                                                                                                                      N
                                                                                                                                                       i




                                                                                                                                                       60
                                                                                                                                                           '




                                                                                                                                                   ImageSotunt




                                                                                                                                                      160
                                                                                                                                                               w
                                                                                                                                                                *u
                                                                                                                                                                     mA
                                                                                                                                                                      .*




                                                                                                                                                     80 100 320 140 160 180
                                                                                                                                                  Image.Eoumt

                                                                                                       Fig. 9 Scatter plot, minutiae count vs. sample number using Aware

                                                                                                      Image quality is another metric that was considered
                                                                                                   important for this research. Fig. 10 shows the scatter plot of
                                                                                                   image quality scores obtained using Aware, Inc.'s quality
                                                                                                   algorithm on the live and gelatin fingerprints. The graph of
                                                                                                   image quality scores clearly indicates a degradation of the
                                                                                                   gelatin fingerprint. T-tests of image quality scores between the
                                                                                                   live and gelatin fingerprints showed a statistically significant
                                                                                                   difference. The severe decrease in image quality noticed in the
                                                                                                   repeated use of the gelatin finger indicates that it would be of
                                                                                                   practical use only for the first 10 or so attempts. Table 1 shows
                                                                                                   the results from the t-tests.

                                                                                                            Scatterplot of AWAREjLiveJIQ, AWARE-Gelatn IQ vshnage-Count
                                                                                                                       U.        *




                                                                                                                                      |
                                                                                                                                                  *    *             **




                                                                                                                                                                          *


                                                                                                                                                                          U




                                                                                                                                                               T-TEST RESULTS




                                                                                                   Interestingly, image quality scores might not provide a clear
                                                                                                   indication of a spoofing attempt, because in the initial nine
                                                                                                   samples, there was no statistically significant difference
                                                                                                   between the image quality score means ofthe two groups.
                                                                                                                                                                              _
                                                                                                                                                                              a




                                                                                                                                                                              X-
                                                                                                                                                                                  a




                                                                                                                                                                 100 120 140 160 180


                                                                                                   Fig. 10 Scatter plot, image quality scores using Aware QualityCheck
                                                                                                                                                                  TABLE I
                                                                                                                                                                                       a 5

                                                                                                                                                                                       s
                                                                                                                                                                                      UI




                                                                                                                                                                                      *i


                                                                                                                                                                                      t-&-




                                                                                                                                                                                       Mean
                                                                                                                                                                                       79.22
                                                                                                                                                                                       61.0'
                                                                                                                                                                                       1.88
                                                                                                                                                                                       2.21
                                                                                                                                                                                             U
                                                                                                                                                                                                    l AWARE Live count
                                                                                                                                                                                                 -_U- AWAREfidati count




                                                                                                                                                                                                  -4-




                                                                                                                                                                                                   -U
                                                                                                                                                                                                        AWARE_i.veJq



                                                                                                                                                                                                        AWAREJ.I*anJ




                                                                                                                                                                                                          p-value
                                                                                                                                                                                                          <.05
                                                                                                                                                                                                          <.05




                                                                                                  34


                Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
B. Experiment 2                                                                      A paired t-test test for statistically significant difference was
Using the Aware software, the minutiae count and quality                             conducted on image quality scores. The test showed a p-value
scores were computed for all the live, and silicon fingerprints                      of < .05, which indicated that the difference in iimage quality
collected in the second experiment. Table II shows the                               scores was statistically significant. The minutiae count for the
summary statistics for image quality and minutiae count for                          dataset of silicon fingerprints and dataset of live fingerprints is
the dataset of live and silicon fingerprints.                                        very similar, but the quality scores for the two groups were
                                                                                     significantly different. The results from this experiment
                                        TABLE II                                     indicate that image quality scores from the silicon fingerprints
                                 SUMMARY STATISTICS                                  are of medium quality, but they still are significantly different
Groups                   N           Median              Median Minutiae             from image quality of live fingers. This provides an
                                     Image               Count                       interesting observation - the extraction of minutiae points is
                     Quality Score                                                   very similar between silicon mold fngerprints and live
Live Fingers    3600 80.0                                39.0                        fingerprint images, but the difference in quality scores points
Silicon Fingers 3600 69.0                                39.0                        to noise in the ridge flow, and contrast levels between ridges
Groups          N    Mean Image                          Mean Minutiae               and valleys and other factors. This could also be possible due
                     Quality Score                       Count                       to distortion and elastic deformation of the silicon fingerprint
Live Fingers    3600 76.46                               39.44                       being different than the corresponding live fingerprint. This
Silicon Fingers 3600 61.35                               38.98                       observation merits ftuther research into this phenomenon since
                                                                                     the ridge flow and contrast analysis could be used as a
  Fig. 11 and 12 show the box plot of quality scores and                             differentiating factor. The spread of the image quality scores
minutiae count respectively. The spread of the image quality                         of the silicon mold is also more variable than the images
scores is a lot larger for silicone finger compared to live                          quality scores of the live fmgerprint images, which indicates
fingers.                                                                             degradation of the silicon fingerprints between successive
                                                                                     attempts.
         Boxplot of Live Finger & Silicone Finger Image Quality Scores                                           VI. CONCLUSION
       inn-I
       iw
                                                                                        The danger with providing a recipe for spoofing is that an
        80
                                                                                    attack methodology to a biometric sensor is revealed.
                                                                                    However, in this case, the attack is analogous to an individual
                                                                                    revealing a PIN number to a fraudster and accompanying the
                                                                                    fraudster to the ATM to watch the fraudster withdraw the
  ii    240
                                                                                     individual's money. The test was not designed as a spoofing
                                                                                    enquiry to evaluate the security of the system, but rather to
                                                                                    understand the characteristics of the gelatin finger and silicon
        20Q
                                                                                    fmger compared to its live counterpart.
                                                                                        Some interesting results were obtained as result of the
         0
                                                                                    analysis in both the experiments. In the first experiment, the
                       Live Finger                    Silio Finger
                                                                                    gelatin finger was able to provide 163 samples with an
 Fig. 11 Box plot, image quality scores using Aware image quality
                                                                                    acquisition rate of 90.7%. Further analysis of fingerprints from
                                        software                                    the live and gelatin fingers showed a considerable difference
                                                                                    in the basic characteristics between the two groups. Repeated
                                                                                    use of the gelatin finger resulted in a rapid degradation of the
                     Boxplot of Live Finger, Silicone Finger                        quality of prints provided, which was reinforced by an
       100-                  *                                                      increase in minutiae count with repeated use. Expecting a
                                                                                    gelatin finger to survive over a 100 attempts would be
        80S                                                                         unreasonable, but our analysis showed that even after 10 uses,
                                                                                    the gelatin finger showed a severe degradation in quality, even
        60-                                                                         though the system matched the spoof. Stabilization of the
  0
                                                                                    minutiae count for the live fingerprint was an unexpected
  A     40-                                                                         result of the experiment, but it reaffins the notion of
                                                                                    habituation and how it can be used to acquire fingerprint
        20-                                                                          samples representative of its owner. Comparing the minutiae
                                                                                     count results of the first and second experiment, the minutiae
         O-                                                                          count of silicon mold fingerprints was more similar to the live
                    Live Mintae Count              Silicone Minutae Count            fingerprints compared to the gelatin mold fingerprints. A
                                                                                     future work of this study is to perform matching operations
                      Fig. 12 Box plot, Minutiae count                               between the live fingerprints and silicon fingerprints to
                                                                                    examine an effect of quality and minutiae count on the
                                                                                   35

               Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
matching error rates. Results from both the experiments
showed a statistically significant difference in image quality
scores between the fake fingerprints and live fingerprints
which could be used as an anti-spoofing mechanism.
Understanding the characteristics of fake fingerprints is
important in devising countermeasures, and extremely
important in increasing security of fingerprint biometric
systems.

                                 REFERENCES
[11 T. Matsumoto, H. Matsumoto, K. Yamada, and S. Hoshino, "Impact of
       artificial 'gummy' fingers on fingerprint systems, i Proc. SPIE, vol.
       4677, Optical Security and Counterfeit Deterrence Techniques IV, San
       Jose, CA, 2002, pp. 275-289.
2]     R. K. Rowe. "A multispectral sensor for fingerprint spoof detection.
       Sensors, " January 2005, p. 4.
[3]    N. K. Ratha, J. H. Connell, and R. M. Bolle, "Enhancing security and
       privacy in biometrics-based authentication systems," IBM Syst J 40(3),
       2001, pp. 614-634.
[4]    A. Mansfield and J. Wayman, "Best practices in testing and reporting
       perfonnances of biometric devices," UK Biometric Working Group,
       City, ST, 2002, pp. 1-32.
[5]    U.S. Department of Homeland Security, Transportation worker
       identification credential (TWIC) program, Available online:
       https://www.twicprogram.com.
[6]    National Institute of Standards and Technology, Personal identity
       verification of federal employees and eontractors, Availabke online:
       http: /csrc.nist.gov/piv-program.
[7]    S. J. Elliott, N. Sickler, and E. Kukula. Automatic identification and data
       capture. 3rd ed., West Lafayette, IN: Copymat. 2005, p. 314.
[8]    A. Jam and N. Duta. "Deformable matching of hand shapes for user
       verification," in 1999 Int Conf Image Processing, 1999, Kobe, Japan:
       IEEE.
[9]    N. Haas, S. Pankanti, and M. Yao, "Fingerprint quality assessment"
       Automatic fingerprint recognition systems, NY: Springer-Verlag, 2004,
       pp. 55-66.
[10]   F. Fernandez, J. Aguilar, and J. Garcia, "A review of schemes for
       fingerprint image quality computation," in 3rd COST 275 Workshop,
       Biometrics on the Internet, Hatfield, UK, 2005.




                                                                                     36

                Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.

More Related Content

What's hot

A novel approach to generate face biometric template using binary discriminat...
A novel approach to generate face biometric template using binary discriminat...A novel approach to generate face biometric template using binary discriminat...
A novel approach to generate face biometric template using binary discriminat...sipij
 
7 multi biometric fake detection system using image quality based liveness de...
7 multi biometric fake detection system using image quality based liveness de...7 multi biometric fake detection system using image quality based liveness de...
7 multi biometric fake detection system using image quality based liveness de...INFOGAIN PUBLICATION
 
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...International Center for Biometric Research
 
An Approach to Speech and Iris based Multimodal Biometric System
An Approach to Speech and Iris based Multimodal Biometric SystemAn Approach to Speech and Iris based Multimodal Biometric System
An Approach to Speech and Iris based Multimodal Biometric SystemIJEEE
 
Role of fuzzy in multimodal biometrics system
Role of fuzzy in multimodal biometrics systemRole of fuzzy in multimodal biometrics system
Role of fuzzy in multimodal biometrics systemKishor Singh
 
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction ModelInternational Center for Biometric Research
 
Feature Level Fusion of Multibiometric Cryptosystem in Distributed System
Feature Level Fusion of Multibiometric Cryptosystem in Distributed SystemFeature Level Fusion of Multibiometric Cryptosystem in Distributed System
Feature Level Fusion of Multibiometric Cryptosystem in Distributed SystemIJMER
 
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...CSCJournals
 
An Investigation towards Effectiveness of Present State of Biometric-Based Au...
An Investigation towards Effectiveness of Present State of Biometric-Based Au...An Investigation towards Effectiveness of Present State of Biometric-Based Au...
An Investigation towards Effectiveness of Present State of Biometric-Based Au...RSIS International
 
Introduction to biometric systems security
Introduction to biometric systems securityIntroduction to biometric systems security
Introduction to biometric systems securitySelf
 
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...International Center for Biometric Research
 
final year embedded system projects in chennai
final year embedded system projects in chennai final year embedded system projects in chennai
final year embedded system projects in chennai Ashok Kumar.k
 
Multi modal biometric system
Multi modal biometric systemMulti modal biometric system
Multi modal biometric systemAalaa Khattab
 
IRJET- A Review on Fake Biometry Detection
IRJET- A Review on Fake Biometry DetectionIRJET- A Review on Fake Biometry Detection
IRJET- A Review on Fake Biometry DetectionIRJET Journal
 
Fake Multi Biometric Detection using Image Quality Assessment
Fake Multi Biometric Detection using Image Quality AssessmentFake Multi Biometric Detection using Image Quality Assessment
Fake Multi Biometric Detection using Image Quality Assessmentijsrd.com
 
OVERVIEW OF MULTIBIOMETRIC SYSTEMS
OVERVIEW OF MULTIBIOMETRIC SYSTEMSOVERVIEW OF MULTIBIOMETRIC SYSTEMS
OVERVIEW OF MULTIBIOMETRIC SYSTEMSAM Publications
 

What's hot (19)

A novel approach to generate face biometric template using binary discriminat...
A novel approach to generate face biometric template using binary discriminat...A novel approach to generate face biometric template using binary discriminat...
A novel approach to generate face biometric template using binary discriminat...
 
7 multi biometric fake detection system using image quality based liveness de...
7 multi biometric fake detection system using image quality based liveness de...7 multi biometric fake detection system using image quality based liveness de...
7 multi biometric fake detection system using image quality based liveness de...
 
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...
(2009) Human-Biometric Sensor Interaction: Impact of Training on Biometric Sy...
 
An Approach to Speech and Iris based Multimodal Biometric System
An Approach to Speech and Iris based Multimodal Biometric SystemAn Approach to Speech and Iris based Multimodal Biometric System
An Approach to Speech and Iris based Multimodal Biometric System
 
IJET-V3I1P25
IJET-V3I1P25IJET-V3I1P25
IJET-V3I1P25
 
Role of fuzzy in multimodal biometrics system
Role of fuzzy in multimodal biometrics systemRole of fuzzy in multimodal biometrics system
Role of fuzzy in multimodal biometrics system
 
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model
(2009) A Definitional Framework for the Human-Biometric Sensor Interaction Model
 
(2008) Impact of Gender on Fingerprint Recognition Systems
(2008) Impact of Gender on Fingerprint Recognition Systems(2008) Impact of Gender on Fingerprint Recognition Systems
(2008) Impact of Gender on Fingerprint Recognition Systems
 
Feature Level Fusion of Multibiometric Cryptosystem in Distributed System
Feature Level Fusion of Multibiometric Cryptosystem in Distributed SystemFeature Level Fusion of Multibiometric Cryptosystem in Distributed System
Feature Level Fusion of Multibiometric Cryptosystem in Distributed System
 
699 703
699 703699 703
699 703
 
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
Robust Analysis of Multibiometric Fusion Versus Ensemble Learning Schemes: A ...
 
An Investigation towards Effectiveness of Present State of Biometric-Based Au...
An Investigation towards Effectiveness of Present State of Biometric-Based Au...An Investigation towards Effectiveness of Present State of Biometric-Based Au...
An Investigation towards Effectiveness of Present State of Biometric-Based Au...
 
Introduction to biometric systems security
Introduction to biometric systems securityIntroduction to biometric systems security
Introduction to biometric systems security
 
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...
(2009) A Comparison of Fingerprint Image Quality and Matching Performance bet...
 
final year embedded system projects in chennai
final year embedded system projects in chennai final year embedded system projects in chennai
final year embedded system projects in chennai
 
Multi modal biometric system
Multi modal biometric systemMulti modal biometric system
Multi modal biometric system
 
IRJET- A Review on Fake Biometry Detection
IRJET- A Review on Fake Biometry DetectionIRJET- A Review on Fake Biometry Detection
IRJET- A Review on Fake Biometry Detection
 
Fake Multi Biometric Detection using Image Quality Assessment
Fake Multi Biometric Detection using Image Quality AssessmentFake Multi Biometric Detection using Image Quality Assessment
Fake Multi Biometric Detection using Image Quality Assessment
 
OVERVIEW OF MULTIBIOMETRIC SYSTEMS
OVERVIEW OF MULTIBIOMETRIC SYSTEMSOVERVIEW OF MULTIBIOMETRIC SYSTEMS
OVERVIEW OF MULTIBIOMETRIC SYSTEMS
 

Viewers also liked (7)

(2003) Securing the Biometric Model
(2003) Securing the Biometric Model(2003) Securing the Biometric Model
(2003) Securing the Biometric Model
 
(2008) Statistical Analysis Framework for Biometric System Interoperability T...
(2008) Statistical Analysis Framework for Biometric System Interoperability T...(2008) Statistical Analysis Framework for Biometric System Interoperability T...
(2008) Statistical Analysis Framework for Biometric System Interoperability T...
 
(2003) Securing a Restricted Site - Biometric Authentication at Entry Point
(2003) Securing a Restricted Site - Biometric Authentication at Entry Point(2003) Securing a Restricted Site - Biometric Authentication at Entry Point
(2003) Securing a Restricted Site - Biometric Authentication at Entry Point
 
Biometric Courses
Biometric CoursesBiometric Courses
Biometric Courses
 
(2011) Face Image Quality
(2011) Face Image Quality(2011) Face Image Quality
(2011) Face Image Quality
 
(2010) Fingerprint Force paper
(2010) Fingerprint Force paper(2010) Fingerprint Force paper
(2010) Fingerprint Force paper
 
(2004) Adaptation and Implementation to a Graduate Course Development in Biom...
(2004) Adaptation and Implementation to a Graduate Course Development in Biom...(2004) Adaptation and Implementation to a Graduate Course Development in Biom...
(2004) Adaptation and Implementation to a Graduate Course Development in Biom...
 

Similar to (2007) Image Quality and Minutiae Count Comparison for Genuine and Artificial Fingerprints

Biometric System ‎Concepts and Attacks
Biometric System ‎Concepts and AttacksBiometric System ‎Concepts and Attacks
Biometric System ‎Concepts and AttacksSaif Salah
 
Intelligent multimodal identification system based on local feature fusion be...
Intelligent multimodal identification system based on local feature fusion be...Intelligent multimodal identification system based on local feature fusion be...
Intelligent multimodal identification system based on local feature fusion be...nooriasukmaningtyas
 
Jss academy of technical education
Jss academy of technical educationJss academy of technical education
Jss academy of technical educationArhind Gautam
 
Fingerprint detection
Fingerprint detectionFingerprint detection
Fingerprint detectionMudit Mishra
 
Biometric Systems and Security
Biometric Systems and SecurityBiometric Systems and Security
Biometric Systems and SecurityShreyans Jain
 
International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)IJERD Editor
 
Antispoofing in face biometrics: A comprehensive study on software-based tech...
Antispoofing in face biometrics: A comprehensive study on software-based tech...Antispoofing in face biometrics: A comprehensive study on software-based tech...
Antispoofing in face biometrics: A comprehensive study on software-based tech...CSITiaesprime
 
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...International Center for Biometric Research
 
Final Report Biometrics
Final Report BiometricsFinal Report Biometrics
Final Report Biometricsanoop80686
 
Security Issues Related to Biometrics
Security Issues Related to BiometricsSecurity Issues Related to Biometrics
Security Issues Related to BiometricsYogeshIJTSRD
 
J018127176.publishing paper of mamatha (1)
J018127176.publishing paper of mamatha (1)J018127176.publishing paper of mamatha (1)
J018127176.publishing paper of mamatha (1)IOSR Journals
 
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...IJCSEIT Journal
 
IRJET- A Review on Security Attacks in Biometric Authentication Systems
IRJET- A Review on Security Attacks in Biometric Authentication SystemsIRJET- A Review on Security Attacks in Biometric Authentication Systems
IRJET- A Review on Security Attacks in Biometric Authentication SystemsIRJET Journal
 

Similar to (2007) Image Quality and Minutiae Count Comparison for Genuine and Artificial Fingerprints (20)

Biometric System ‎Concepts and Attacks
Biometric System ‎Concepts and AttacksBiometric System ‎Concepts and Attacks
Biometric System ‎Concepts and Attacks
 
Biometricsppt
BiometricspptBiometricsppt
Biometricsppt
 
Cover page
Cover pageCover page
Cover page
 
(2009) Statistical Analysis Of Fingerprint Sensor Interoperability
(2009) Statistical Analysis Of Fingerprint Sensor Interoperability(2009) Statistical Analysis Of Fingerprint Sensor Interoperability
(2009) Statistical Analysis Of Fingerprint Sensor Interoperability
 
Intelligent multimodal identification system based on local feature fusion be...
Intelligent multimodal identification system based on local feature fusion be...Intelligent multimodal identification system based on local feature fusion be...
Intelligent multimodal identification system based on local feature fusion be...
 
Jss academy of technical education
Jss academy of technical educationJss academy of technical education
Jss academy of technical education
 
Fingerprint detection
Fingerprint detectionFingerprint detection
Fingerprint detection
 
Biometric Systems and Security
Biometric Systems and SecurityBiometric Systems and Security
Biometric Systems and Security
 
International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)
 
Antispoofing in face biometrics: A comprehensive study on software-based tech...
Antispoofing in face biometrics: A comprehensive study on software-based tech...Antispoofing in face biometrics: A comprehensive study on software-based tech...
Antispoofing in face biometrics: A comprehensive study on software-based tech...
 
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...
(2005) An Evaluation Of Fingerprint Image Quality Across An Elderly Populatio...
 
Final Report Biometrics
Final Report BiometricsFinal Report Biometrics
Final Report Biometrics
 
BIOMETRIC SECURITY SYSTEM AND ITS APPLICATIONS IN HEALTHCARE
BIOMETRIC SECURITY SYSTEM AND ITS APPLICATIONS IN HEALTHCAREBIOMETRIC SECURITY SYSTEM AND ITS APPLICATIONS IN HEALTHCARE
BIOMETRIC SECURITY SYSTEM AND ITS APPLICATIONS IN HEALTHCARE
 
Biometric technologies
Biometric technologies Biometric technologies
Biometric technologies
 
Security Issues Related to Biometrics
Security Issues Related to BiometricsSecurity Issues Related to Biometrics
Security Issues Related to Biometrics
 
J018127176.publishing paper of mamatha (1)
J018127176.publishing paper of mamatha (1)J018127176.publishing paper of mamatha (1)
J018127176.publishing paper of mamatha (1)
 
(2007) Performance Analysis for Multi Sensor Fingerprint Recognition System
(2007) Performance Analysis for Multi Sensor Fingerprint Recognition System(2007) Performance Analysis for Multi Sensor Fingerprint Recognition System
(2007) Performance Analysis for Multi Sensor Fingerprint Recognition System
 
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...
BIOMETRICS AUTHENTICATION TECHNIQUE FOR INTRUSION DETECTION SYSTEMS USING FIN...
 
Iciea08
Iciea08Iciea08
Iciea08
 
IRJET- A Review on Security Attacks in Biometric Authentication Systems
IRJET- A Review on Security Attacks in Biometric Authentication SystemsIRJET- A Review on Security Attacks in Biometric Authentication Systems
IRJET- A Review on Security Attacks in Biometric Authentication Systems
 

More from International Center for Biometric Research

An Investigation into Biometric Signature Capture Device Performance and User...
An Investigation into Biometric Signature Capture Device Performance and User...An Investigation into Biometric Signature Capture Device Performance and User...
An Investigation into Biometric Signature Capture Device Performance and User...International Center for Biometric Research
 
Advances in testing and evaluation using Human-Biometric sensor interaction m...
Advances in testing and evaluation using Human-Biometric sensor interaction m...Advances in testing and evaluation using Human-Biometric sensor interaction m...
Advances in testing and evaluation using Human-Biometric sensor interaction m...International Center for Biometric Research
 
(2010) Fingerprint recognition performance evaluation for mobile ID applications
(2010) Fingerprint recognition performance evaluation for mobile ID applications(2010) Fingerprint recognition performance evaluation for mobile ID applications
(2010) Fingerprint recognition performance evaluation for mobile ID applicationsInternational Center for Biometric Research
 

More from International Center for Biometric Research (20)

HBSI Automation Using the Kinect
HBSI Automation Using the KinectHBSI Automation Using the Kinect
HBSI Automation Using the Kinect
 
IT 34500
IT 34500IT 34500
IT 34500
 
An Investigation into Biometric Signature Capture Device Performance and User...
An Investigation into Biometric Signature Capture Device Performance and User...An Investigation into Biometric Signature Capture Device Performance and User...
An Investigation into Biometric Signature Capture Device Performance and User...
 
Entropy of Fingerprints
Entropy of FingerprintsEntropy of Fingerprints
Entropy of Fingerprints
 
Biometric and usability
Biometric and usabilityBiometric and usability
Biometric and usability
 
Examining Intra-Visit Iris Stability - Visit 4
Examining Intra-Visit Iris Stability - Visit 4Examining Intra-Visit Iris Stability - Visit 4
Examining Intra-Visit Iris Stability - Visit 4
 
Examining Intra-Visit Iris Stability - Visit 6
Examining Intra-Visit Iris Stability - Visit 6Examining Intra-Visit Iris Stability - Visit 6
Examining Intra-Visit Iris Stability - Visit 6
 
Examining Intra-Visit Iris Stability - Visit 2
Examining Intra-Visit Iris Stability - Visit 2Examining Intra-Visit Iris Stability - Visit 2
Examining Intra-Visit Iris Stability - Visit 2
 
Examining Intra-Visit Iris Stability - Visit 1
Examining Intra-Visit Iris Stability - Visit 1Examining Intra-Visit Iris Stability - Visit 1
Examining Intra-Visit Iris Stability - Visit 1
 
Examining Intra-Visit Iris Stability - Visit 3
Examining Intra-Visit Iris Stability - Visit 3Examining Intra-Visit Iris Stability - Visit 3
Examining Intra-Visit Iris Stability - Visit 3
 
Best Practices in Reporting Time Duration in Biometrics
Best Practices in Reporting Time Duration in BiometricsBest Practices in Reporting Time Duration in Biometrics
Best Practices in Reporting Time Duration in Biometrics
 
Examining Intra-Visit Iris Stability - Visit 5
Examining Intra-Visit Iris Stability - Visit 5Examining Intra-Visit Iris Stability - Visit 5
Examining Intra-Visit Iris Stability - Visit 5
 
Standards and Academia
Standards and AcademiaStandards and Academia
Standards and Academia
 
Interoperability and the Stability Score Index
Interoperability and the Stability Score IndexInteroperability and the Stability Score Index
Interoperability and the Stability Score Index
 
Advances in testing and evaluation using Human-Biometric sensor interaction m...
Advances in testing and evaluation using Human-Biometric sensor interaction m...Advances in testing and evaluation using Human-Biometric sensor interaction m...
Advances in testing and evaluation using Human-Biometric sensor interaction m...
 
Cerias talk on testing and evaluation
Cerias talk on testing and evaluationCerias talk on testing and evaluation
Cerias talk on testing and evaluation
 
IT 54500 overview
IT 54500 overviewIT 54500 overview
IT 54500 overview
 
Ben thesis slideshow
Ben thesis slideshowBen thesis slideshow
Ben thesis slideshow
 
(2010) Fingerprint recognition performance evaluation for mobile ID applications
(2010) Fingerprint recognition performance evaluation for mobile ID applications(2010) Fingerprint recognition performance evaluation for mobile ID applications
(2010) Fingerprint recognition performance evaluation for mobile ID applications
 
ICBR Databases
ICBR DatabasesICBR Databases
ICBR Databases
 

Recently uploaded

9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding TeamAdam Moalla
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaborationbruanjhuli
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...Aggregage
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UbiTrack UK
 
Introduction to Matsuo Laboratory (ENG).pptx
Introduction to Matsuo Laboratory (ENG).pptxIntroduction to Matsuo Laboratory (ENG).pptx
Introduction to Matsuo Laboratory (ENG).pptxMatsuo Lab
 
Bird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemBird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemAsko Soukka
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureEric D. Schabell
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?IES VE
 
Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024D Cloud Solutions
 
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdf
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdf
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdfJamie (Taka) Wang
 
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfUiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfDianaGray10
 
NIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopNIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopBachir Benyammi
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdfPedro Manuel
 
Computer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsComputer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsSeth Reyes
 
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Will Schroeder
 
Cybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxCybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxGDSC PJATK
 
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsIgniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsSafe Software
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfinfogdgmi
 
AI You Can Trust - Ensuring Success with Data Integrity Webinar
AI You Can Trust - Ensuring Success with Data Integrity WebinarAI You Can Trust - Ensuring Success with Data Integrity Webinar
AI You Can Trust - Ensuring Success with Data Integrity WebinarPrecisely
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6DianaGray10
 

Recently uploaded (20)

9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
 
Introduction to Matsuo Laboratory (ENG).pptx
Introduction to Matsuo Laboratory (ENG).pptxIntroduction to Matsuo Laboratory (ENG).pptx
Introduction to Matsuo Laboratory (ENG).pptx
 
Bird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemBird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystem
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability Adventure
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?
 
Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024
 
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdf
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdf
activity_diagram_combine_v4_20190827.pdfactivity_diagram_combine_v4_20190827.pdf
 
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdfUiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
UiPath Solutions Management Preview - Northern CA Chapter - March 22.pdf
 
NIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopNIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 Workshop
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdf
 
Computer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsComputer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and Hazards
 
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
 
Cybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxCybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptx
 
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration WorkflowsIgniting Next Level Productivity with AI-Infused Data Integration Workflows
Igniting Next Level Productivity with AI-Infused Data Integration Workflows
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdf
 
AI You Can Trust - Ensuring Success with Data Integrity Webinar
AI You Can Trust - Ensuring Success with Data Integrity WebinarAI You Can Trust - Ensuring Success with Data Integrity Webinar
AI You Can Trust - Ensuring Success with Data Integrity Webinar
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6
 

(2007) Image Quality and Minutiae Count Comparison for Genuine and Artificial Fingerprints

  • 1. Image Quality and Minutiae Count Comparison for Genuine and Artificial Fingerprints Stephen J. Elliott Ph.D., Shimon K. Modi, Lou Maccarone, Matthew R. Young, Changlong Jin, and Hale Kim Ph.D. same user after acquisition on a commercially available Abstract-The vulnerabilities of biometric sensors have been biometric fingerprint sensor. This research seeks to answer discussed extensively in the literature and popularized in films and whether the samples (live versus gelatin) exhibit the same television shows. This research examines the image quality of an minutiae counts and whether the acquired images possess the artificial print as compared to a genuine finger, and examines the same image quality properties. That is, are the live and gelatin characteristics of the two, including minutiae counts and image quality, as repeated samples are taken. samples similar in their characteristics, and do their similarities or differences have an impact on matching Keywords-authentication, biometrics, fingerprint recognition, perfonnance? repeatability. II. BIOMETRIC SYSTEM VULNERABILITIES I. INTRODUCTION AND MOTIVATION All security measures, including mechanisms for V ERIFYING the identity of an individual can be authenticating identity, can be circumvented. The processes Vaccomplished by three main considerations: what an associated with working around these measures vary in individual has, what an individual knows or owns, and what difficulty according to the level of effort and resources needed an individual is. The first option is typically achieved through to cany out the deceptive act. Authentication mechanisms the use of a token, such as an identification card, badge, based on secrets are particularly vulnerable to "guessing" magnetic stripe, or radio frequency identification (RFID) tag. attacks. Token mechanisms that rely on the possession of an The second option can be achieved through the use of a object, typically a card or badge technology are most password or personal identification number (PIN). The third vulnerable to theft or falsified reproduction. Biometric option can be accomplished through what an individual is by technologies closely tie the authenticator to the individual utilizing biometric technologies. Like the first two identity of the user through the use of physiological or authentication methods, biometric systems contain behavioral characteristics. While this property offers an added vulnerabilities and are susceptible to attack. Some of these advantage over the other two authentication mechanisms, it vulnerabilities are similar or even overlapping across all three places a great emphasis on validating the integrity of the authentication mechanisms. However, attacks specific to biometric sample acquired and transferred in the biometric biometric systems focus on liveness detection of a human (i.e., system. Ratha, Connell, and Bolle provided a model whether the finger from a live sample or a gelatin sample). identifying vulnerabilities in biometric systems [3]. An Various attacks documented in the literature have focused on example of the threat model is shown in Fig. 1, and builds on the sensor [11, [2]. While understanding and preventing the general biometric model outlined in Mansfield and attacks on the sensor is an interesting research topic worthy of Wayman [4]. investigation, this paper examines the global and local features of a live sample compared to that of a gelatin finger from the S. J. Elliott, Ph.D. is an Associate Professor with the Departnent of Industrial Technology at Purdue University, West Lafayette, IN 47907 USA; (e-mail:elliott apurdue.edu . S. Modi is graduate student with the Departnent of Industrial Technology at Purdue University, West Lafayette, IN 47907 USA; (e-mail:shimon@ purdue.edu). L. Maccarone is an undergraduate student at Purdue University, West Lafayette, IN 47907 USA Fig. 1 Biometric system threat model M. R. Young is a graduate student with the Department of Industrial Technology at Purdue University, West Lafayette, IN 47907 USA C. Jin is a graduate student with the School of Information and The biometric system threat model shown in Fig. 1 contains Communication Engineering at Inha University, Incheon, Korea; (e- 11 individual areas of vulnerability. In addition to the five mail:cljin@vision.inha.ac.kr). main internal modules characterized in the general biometric H. Kim is a Professor with the School of Infornation and Communication Engineering at Inha University, Incheon, Korea; (e-mail:hikim@inha.ac.kr). model (data collection, signal processing, matching, storage, 1-4244-1129-7/07/$25.00 ©2007 IEEE 30 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 2. and decision) [4], another component is added to represent the the features of the two fingerprints remain consistent transfer of the authentication decision to the application that Repeatability of the extracted features is important for the relies on the decision from the, biometric system. Such matching process in any type of biometric technology [7]. The applications could be identity management systems (IDMS) or features to be examined include: minutiae points and image access control systems for logical and or physical access to quality. One of the challenges associated with this research is resources. These systems can vary in complexity and size, to ensure the image is of sufficient quality. A wide variety of ranging from a local computer log-in all the way to a wide- factors can influence the quality of fingerprint samples. Non- scale distributed architecture seen in the cases of the U.S. uniform contact, inconsistent contact, or irreproducible contact Department of Transportation's Transportation Worker with the fingerprint sensor can result in images with a low Identification Credential (TWIC) [5] or the Personal Identity signal-to-noise ratio, which is not desirable for feature Verification (PIV) of Federal Employees and Contractors [6]. extraction and matching purposes [9]. Wear and tear of the The remaining points of vulnerability are communication skin and aging effects can semi-permanently alter ridge channels between these six modules. It is worth noting that not characteristics. These factors also affect acquisition of all 11 vulnerability points are unique to the biometric system. fingerprints by the fingerprint sensor. The importance of Many of the same points (e.g., storage and communication quality is widely acknowledged, but there is no standard channels) are vulnerable in other authentication systems and means of assessing quality. The current standardization effort similar methods can be used to limit those particular for assessing quality for biometric samples refers to three vulnerabilities. different connotations of quality: The most publicized vulnerability in biometric systems * Character resides at the data collection module in the form of spoofmg or * Fidelity presenting artificial representations of biometric samples (see * Utility module #1, Fig. 1). If an artificial or fake biometric sample is These three connotations of biometric sample quality can be accepted by the biometric system at this initial stage, the entire directly applied to fingerprint sample quality. Character is a biometric system is corrupted and the system is compromised. description of quality based on inherent features from the Attacks on biometric systems are not new, popular culture source of the fingerprint. Individuals who have scarred seeks to circumvent security systems and biometric systems fingerprints or dry or cracked skin on the fingertips will are not immune. Several online resources are available that provide samples with poor character. Fidelity is a description describe such attacks on the data collection module, and many of quality based on degree of similarity between the actual movies and television shows have highlighted attacks on such fingerprint and the fingerprint image acquired by the sensor. systems. One such attack at this data collection module was Inconsistent contact with the fmgerprint sensor can lead to outlined in the work of Matsumoto, Yamada, and Hoshino fingerprint samples with poor fidelity. Utility is a description (2002) using "Gummy Fingers" [1]. of quality based on observed or predicted contribution of the The biometric research community, as well as industry, has fingerprint sample to the overall performance of the focused its research on preventing such attacks by using the fingerprint recognition system. Utility of a fmgerprint sample concept of "liveness" detection techniques. Today, the newer is directly affected by the character and fidelity of the sensors are improving their resilience against a spoofing attack fingerprint sample, and should be the closely related to at this module. In the past, acetate spoofing attacks - where performance of the recognition system. an image of a fingerprint placed on acetate was accepted as a A substantial amount ofresearch has been conducted in area genuine live finger - was easy to reproduce. Now, such of quality assessment, all of which give varying levels of attacks are proving increasingly difficult to succeed, hence the importance to character, fidelity, and utility. Previous research more complicated approaches to attacks being waged on the in the field of fingerprint image quality assessment can be vulnerable sensor. Techniques for liveness detection within generalized into three categories: local features analysis, the fingerprint modality focus on moisture content, global features analysis, and quality analysis as a classification temperature, electrical conductivity, and challenge response. problem [10]. Features of the fingerprint image such as minutiae count, fidelity of minutiae, contrast ratio between III. FINGERPRnT IMAGE QUALITY ANALYSIS ridges and valleys, capture area of the figerprint, and The purpose of this research paper was not to prove the detennination of dominant direction are used by quality vulnerability of the biometric system, but to examine the algorithms in varying capacities to make quality repeatability of the features of the gelatin finger print as determinations. For the purpose of this research, fingerprints compared to the live genuine sample once the image has been from live fngers and gelatin fingers were examined by two acquired. The research question is whether an artificial print different image quality algorithms, one provided by Aware, captured on an optical sensor exhibits any of the same Inc. and the other provided as a part of NIST Fingerprint characteristics as a genuine fingerprint from the same Imaging Software (NFIS). individual captured on the same sensor, and whether any distinguishers might enable the artificial print to be excluded IV. METHODOLOGY later on in the process, if the initial data collection module This research involved two separate experiments. The first accepts it. The research also investigates whether, over time, 31 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 3. experiment was conducted in two stages. Stage 1 of the first verified by the software, then the mold would have been experiment involved creation of a set of images from an destroyed and the process started again. artificial gelatin fimger that was crafted from the same subject The artificial finger was returned to the refrigerator in a who provided the genuine fingerprint. The procedures to simple (airtight, but not vacuum-sealed) plastic storage accomplish this feat required adaptation of several different container for 48 hours at a temperature of 2TC. This procedure methodologies outlined in the literature for creating an allowed the gelatin finger to completely solidify to its artificial fingerprints, including the work done by Matsumoto, permanent state. Yamada, and Hoshino [1]. Prior to creating the mold, the After removing the gelatin finger from the refrigerator, tests following necessary ingredients and utensils were gathered: were conducted on it to estimate the optimal load required for plastic clay, hot water, and a pair of plastic tongs. To create acquiring images. In general terms, it is best to use the least the mold, a quantity of plastic clay sufficient to cover the weight possible to produce a scan in order to minimize the genuine finger was required. In order to make the plastic clay spreading and dissolution of the gelatin finger's valleys and malleable, it was placed briefly in boiling water. In order to be ridges. Testing of loads ranging from 20Og to 1,000g, as utilized, the plastic clay had to have a consistency such that it measured by a Tanita digital scale, was performed. enabled a mold to be created by placing the finger with only Approximately 200g was determined to be the lower limit to light pressure. When the plastic clay has attained a sufficient produce an image, with 550g being the upper limit before level of pliability, the clay was cooled. Once the clay had distortion and inability to match occurred. The next stage of cooled enough to be touched, the finger was placed into the the experiment involved acquisition of a series of images from the gelatin finger. All of the images were acquired from the clay to a depth sufficient to create the mold to be used to craft optical sensor using the gelatin fimger over a 15-minute time the gelatin finger. The genuine finger was kept in the plastic period. After 15 minutes of acquiring images, the gelatin clay until the clay had cooled enough to retain its shape and finger had degraded to the point at which it was no longer able the details of the genuine finger. After the finger was to be accepted by the optical sensor. In all, 163 images were removed, the mold was allowed to cure for an additional 10 produced over the 15-minute time period. A detailed minutes. The resulting mold for this study is shown in Fig. 2. description of the 160 images utilized in the study is provided in the results section. Stage 2 of the first experiment called for the collection of a series of live samples. One hundred sixty live samples were acquired from the same finger that was used to create the artificial gelatin finger. The live samples were collected over an 8-minute time period on a commercially available optical sensor. The authors chose to collect 160 live samples, as this was the same number of fingerprints collected from the gelatin finger. These 160 live images were all stored according to the time collected; they are used to provide a baseline quality assessment that will be compared against the samples Fig. 2 Plastic mold formed to create gelatin finger generated by the gelatin finger. After data collection, both sets of images (from the live The next step was to create a gelatin mixture capable of finger and the gelatin fimger) were processed through the NIST producing artificial fingerprints from the mold that would be Fingerprint Image Software (NFIS) package. The MINDTCT recognizable to the sensor. Two sheets of gelatin weighing function was used to count the number of minutiae present in 3.5g were soaked in cold water for five minutes. In order to each individual image. The NFIQ function was used to remove the excess water, the gelatin sheets were dried until evaluate image quality, which is determined on a rating scale the gelatin weighed 14-16g. Next, a bowl was immersed into of 1 to 5, with 1 being the best and 5 being the worst. The extremely hot water, and the gelatin was placed into the bowl results of M1NDTCT and NFIQ from both groups (the live to soften and melt the gelatin. Once the gelatin had melted, it fingerprints and the gelatin fingerprints) were then compared was poured into the clay mold. Immediately after placing the by the means of statistical t-tests (using an a level of 0.05) to gelatin in the mold, the mold was placed in a refrigerator to determine if any statistical difference existed across the cool for 10 minutes at a temperature of 1°C. Cooling groups. Aware, Inc. offers a commercially available image transforms the gelatin to a state that is resistant to changes in quality and minutiae count software; this software was used to shape when touched. Ambient room temperate, when this extract image quality scores and minutiae counts for experiment was conducted, was 220C. fingerprints from the live finger and gelatin finger groups. By After the gelatin finger had cured in the refrigerator for a utilizing two different software packages to analyze the period of approximately one hour, it was removed from the fingerprints, we sought to eliminate bias that might have been clay mold and placed on the sensor to determine whether it generated by utilizing only a single application. was actually able to produce images. Ten attempts to acquire The second experiment involved collecting fingerprint an image were made; in all 10 instances, an image was images from left index finger and the right thumb from 30 produced. The software verified the mold. If the mold was not different subjects. Each subject was asked to provide 20 32 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 4. images of each fimger on three different optical sensors. A of live and gelatin fingerprint minutiae count for last 16 prints. silicon mold was created the left index and right thumb of An interesting observation is that the minutiae count increases each subject, and the molds were used to provide 20 for the gelatin fingerprint group, but stabilizes for live fingerprint images on each of the three optical sensors. The fingerprints. procedures used in [1] were also employed in creating the silicon molds in the second part of the experiment. At the end Boxplot of AWAREJLive-comt, AWARE_Gelatincount of the data collection there were 3600 fingerprint images from 70. live fingers, and the 3600 fingerprint images from silicon fingers. 60* A. Experiment I V. RESULTS a 50I Creation and acquisition of images from gelatin fingers can be problematic, as previous research has shown that gelatin 40 _ fingers do not afford consistent repeatability. However, this study provides anecdotal evidence suggesting that better 3J preparation and storage of the artificial finger can aid in the repeatability of the images produced. The first 39 samples AWARE Live ca,t AWARELGelItkSem_t provided consistently successful spoofing results; on the 40th Fig. 4 Box plot, live and gelatin finger minutiae count using Aware presentation of the gelatin finger, a failure-to-acquire (FTA) QualityCheck resulted. Overall, 163 images were acquired, but only 160 images were used for the final study. Degradation on the final 3 images rendered the images unusable. The acquisition rate Boxplot of t.S_Live_Coumt, NFI&Gelatin cout for this particular gelatin fingerprint was 90.7%, producing a 110 FTA rate of 9.3%. Fig. 3 shows the gelatin print (left) a live print (right). The FTA rate for the live finger was 0.0%. 90- 900 a 0- 60- o 60 50 40 NRSLve_Cort NFS.Geeatincount Fig. 5 Box plot, live and gelatin finger minutiae count using NFIS M1NDTCT Fig. 3 Gelatin finger (left) and live fmger (right) The stabilization of minutiae count for the live fingerprints can probably be attributed to habituation or acclimation to the The minutiae count analysis on both the fingerprint groups device. The subject has been acclimated to placing the sample was performed. Fig. 4 and Fig. 6 show box plots of the finger on the optical sensor, which reduces the inconsistent minutiae count from the live finger and gelatin finger, contact of finger surface with the platen of the sensor. Another respectively, generated from Aware, Inc.'s image quality tool. interesting observation is the increase in minutiae count Fig. 5 and Fig. 7 show box plots of the minutiae count from between the gelatin fingerprints over time. This suggests the live finger and gelatin fmger, respectively, generated from degradation of the gelatin finger and mold because of NFIS's MINDTCT. repetitive use and introduction of cracks in the mold used to The results from the box plot graphs generated by both the create the gelatin finger. Evidence suggests that, over time, the Aware and NFIS software programs show that the live number of minutiae for the gelatin fingerprint increases, while fingerprints have a lower minutiae count than the gelatin the number of minutiae for the live fmgerprint stabilizes. fingerprints, which is most likely a result of indirect and inconsistent contact with the optical sensor. In order to study the deterioration of the gelatin fingerprints, the first 16 and last 16 samples from the live and gelatin fmgerprint groups were used. Fig 6 shows a box plot of the live and gelatin fingerprint minutiae count for first 16 prints, and Fig. 7 shows a box plot 33 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 5. 9 55 501 45 40 35 303 65- So 60- 55- 45- 40- 35- 30 Boxplot of Lisfe_First-16;, Gelatin-First-16 I ve-FirsL6 Boxplot of Live_Last_6, GelatinjLastL16 Live_Last_16 'I'-- Fig. 8 and Fig. 9 shows the scatter plots for minutiae count versus sample numbers of live and gelatin fingerprint groups. Both of these graphs give credence to the observations made from the box plots. Scatterplot of NFISJJve-Count, WISjGelatin_count vs rnageCount > 110 100 80 70 6.0 40 0 2 0 I 1 S%mN& 40 me_pmft U . Jj,#MO 6 U *- 0 " 100 12 1 14 U 1 Fig. 8 Scatter plot, minutiae count vs. sample number using NFIS MIINDTCT Gebtin_Fst16 Fig. 6 Box plot, live and gelatin finger minutiae count using Aware, first 16 prints Gelati_Last_16 Fig. 7 Box plot, live and gelatin finger minutiae count using Aware QualityCheck, last 16 prints - VariabLa -+- EIS Lwe.Co.wA -'- W- NFIS_GeaUn_cwnt (U 'U 9 Groups 70 60- 50- 40- 30- 70 so 80 40 Aware_LiveLIQ 0 Aware_Gelatin_IQ 160 NFIS_Live-IQ 160 NFIS Gelatin IQ 160 * 20 tE a ... Scatterplot of AWAREJive_ount, AWARE_Gelatincmnt vs bnaJeCoLnt ,V,able , 40 5 U0 0406 a 60 * ILI *rn * N i 60 ' ImageSotunt 160 w *u mA .* 80 100 320 140 160 180 Image.Eoumt Fig. 9 Scatter plot, minutiae count vs. sample number using Aware Image quality is another metric that was considered important for this research. Fig. 10 shows the scatter plot of image quality scores obtained using Aware, Inc.'s quality algorithm on the live and gelatin fingerprints. The graph of image quality scores clearly indicates a degradation of the gelatin fingerprint. T-tests of image quality scores between the live and gelatin fingerprints showed a statistically significant difference. The severe decrease in image quality noticed in the repeated use of the gelatin finger indicates that it would be of practical use only for the first 10 or so attempts. Table 1 shows the results from the t-tests. Scatterplot of AWAREjLiveJIQ, AWARE-Gelatn IQ vshnage-Count U. * | * * ** * U T-TEST RESULTS Interestingly, image quality scores might not provide a clear indication of a spoofing attempt, because in the initial nine samples, there was no statistically significant difference between the image quality score means ofthe two groups. _ a X- a 100 120 140 160 180 Fig. 10 Scatter plot, image quality scores using Aware QualityCheck TABLE I a 5 s UI *i t-&- Mean 79.22 61.0' 1.88 2.21 U l AWARE Live count -_U- AWAREfidati count -4- -U AWARE_i.veJq AWAREJ.I*anJ p-value <.05 <.05 34 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 6. B. Experiment 2 A paired t-test test for statistically significant difference was Using the Aware software, the minutiae count and quality conducted on image quality scores. The test showed a p-value scores were computed for all the live, and silicon fingerprints of < .05, which indicated that the difference in iimage quality collected in the second experiment. Table II shows the scores was statistically significant. The minutiae count for the summary statistics for image quality and minutiae count for dataset of silicon fingerprints and dataset of live fingerprints is the dataset of live and silicon fingerprints. very similar, but the quality scores for the two groups were significantly different. The results from this experiment TABLE II indicate that image quality scores from the silicon fingerprints SUMMARY STATISTICS are of medium quality, but they still are significantly different Groups N Median Median Minutiae from image quality of live fingers. This provides an Image Count interesting observation - the extraction of minutiae points is Quality Score very similar between silicon mold fngerprints and live Live Fingers 3600 80.0 39.0 fingerprint images, but the difference in quality scores points Silicon Fingers 3600 69.0 39.0 to noise in the ridge flow, and contrast levels between ridges Groups N Mean Image Mean Minutiae and valleys and other factors. This could also be possible due Quality Score Count to distortion and elastic deformation of the silicon fingerprint Live Fingers 3600 76.46 39.44 being different than the corresponding live fingerprint. This Silicon Fingers 3600 61.35 38.98 observation merits ftuther research into this phenomenon since the ridge flow and contrast analysis could be used as a Fig. 11 and 12 show the box plot of quality scores and differentiating factor. The spread of the image quality scores minutiae count respectively. The spread of the image quality of the silicon mold is also more variable than the images scores is a lot larger for silicone finger compared to live quality scores of the live fmgerprint images, which indicates fingers. degradation of the silicon fingerprints between successive attempts. Boxplot of Live Finger & Silicone Finger Image Quality Scores VI. CONCLUSION inn-I iw The danger with providing a recipe for spoofing is that an 80 attack methodology to a biometric sensor is revealed. However, in this case, the attack is analogous to an individual revealing a PIN number to a fraudster and accompanying the fraudster to the ATM to watch the fraudster withdraw the ii 240 individual's money. The test was not designed as a spoofing enquiry to evaluate the security of the system, but rather to understand the characteristics of the gelatin finger and silicon 20Q fmger compared to its live counterpart. Some interesting results were obtained as result of the 0 analysis in both the experiments. In the first experiment, the Live Finger Silio Finger gelatin finger was able to provide 163 samples with an Fig. 11 Box plot, image quality scores using Aware image quality acquisition rate of 90.7%. Further analysis of fingerprints from software the live and gelatin fingers showed a considerable difference in the basic characteristics between the two groups. Repeated use of the gelatin finger resulted in a rapid degradation of the Boxplot of Live Finger, Silicone Finger quality of prints provided, which was reinforced by an 100- * increase in minutiae count with repeated use. Expecting a gelatin finger to survive over a 100 attempts would be 80S unreasonable, but our analysis showed that even after 10 uses, the gelatin finger showed a severe degradation in quality, even 60- though the system matched the spoof. Stabilization of the 0 minutiae count for the live fingerprint was an unexpected A 40- result of the experiment, but it reaffins the notion of habituation and how it can be used to acquire fingerprint 20- samples representative of its owner. Comparing the minutiae count results of the first and second experiment, the minutiae O- count of silicon mold fingerprints was more similar to the live Live Mintae Count Silicone Minutae Count fingerprints compared to the gelatin mold fingerprints. A future work of this study is to perform matching operations Fig. 12 Box plot, Minutiae count between the live fingerprints and silicon fingerprints to examine an effect of quality and minutiae count on the 35 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.
  • 7. matching error rates. Results from both the experiments showed a statistically significant difference in image quality scores between the fake fingerprints and live fingerprints which could be used as an anti-spoofing mechanism. Understanding the characteristics of fake fingerprints is important in devising countermeasures, and extremely important in increasing security of fingerprint biometric systems. REFERENCES [11 T. Matsumoto, H. Matsumoto, K. Yamada, and S. Hoshino, "Impact of artificial 'gummy' fingers on fingerprint systems, i Proc. SPIE, vol. 4677, Optical Security and Counterfeit Deterrence Techniques IV, San Jose, CA, 2002, pp. 275-289. 2] R. K. Rowe. "A multispectral sensor for fingerprint spoof detection. Sensors, " January 2005, p. 4. [3] N. K. Ratha, J. H. Connell, and R. M. Bolle, "Enhancing security and privacy in biometrics-based authentication systems," IBM Syst J 40(3), 2001, pp. 614-634. [4] A. Mansfield and J. Wayman, "Best practices in testing and reporting perfonnances of biometric devices," UK Biometric Working Group, City, ST, 2002, pp. 1-32. [5] U.S. Department of Homeland Security, Transportation worker identification credential (TWIC) program, Available online: https://www.twicprogram.com. [6] National Institute of Standards and Technology, Personal identity verification of federal employees and eontractors, Availabke online: http: /csrc.nist.gov/piv-program. [7] S. J. Elliott, N. Sickler, and E. Kukula. Automatic identification and data capture. 3rd ed., West Lafayette, IN: Copymat. 2005, p. 314. [8] A. Jam and N. Duta. "Deformable matching of hand shapes for user verification," in 1999 Int Conf Image Processing, 1999, Kobe, Japan: IEEE. [9] N. Haas, S. Pankanti, and M. Yao, "Fingerprint quality assessment" Automatic fingerprint recognition systems, NY: Springer-Verlag, 2004, pp. 55-66. [10] F. Fernandez, J. Aguilar, and J. Garcia, "A review of schemes for fingerprint image quality computation," in 3rd COST 275 Workshop, Biometrics on the Internet, Hatfield, UK, 2005. 36 Authorized licensed use limited to: Purdue University. Downloaded on February 18,2010 at 15:00:20 EST from IEEE Xplore. Restrictions apply.