Presented at The 7th International Conference on Information Technology and Applications (ICITA 2011), Sydney Australia, 21 Nov - 24 Nov 2011.
The purpose of this presentation is to provide additional analysis of image quality and Henry Classification on Finger location on a single sensor. One hundred and sixty nine individuals provided six impressions of their left index, left middle, right index, and right middle fingers. The results show that there is significant difference in image quality, Henry classifications, and zoo animal distribution across the four finger locations under study. The results of this research show that location is an important consideration when developing enrollment best practices for single print systems.
(2011) Image Quality, Performance, and Classification - the Impact of Finger Location
1. BIOMETRICS LAB
Biometric Standards, Performance and Assurance Laboratory
Department of Technology, Leadership and Innovation
IMAGE QUALITY, PERFORMANCE,
AND CLASSIFICATION – THE
IMPACT OF FINGER LOCATION
Purdue University: Michael Brockly | Stephen Elliott
2. RESEARCH QUESTIONS
• How does Henry Classification differ
across finger locations?
• Does image quality differ across finger
locations?
• Does minutiae count differ across finger
locations
• Does finger location impact
performance?
3. RESPONSIVE
• Further the understanding of Henry
Classifications
• Refine zoo plot analysis
• Support ideal finger theories based on
image quality and minutiae
5. SUBJECTS
• Examined a subject pool of 190 users.
• Collected from a multi-sensor study
• Many subjects were missing images due
to error, either data collection or Failure
to Acquire (FTA)
• Reduced the subject pool to 169
subjects to ensure equal numbers of
fingers
6. SUBJECT SUBSET
• 169 subjects
• 118 male, 49 female
• 148 office workers, 16 manual laborers
6154534947434038373635343331302928272625242322212019
40
30
20
10
0
Age
Frequency
User distribution of age
7. SUBJECT SUBSET
• Each subject provided six successful
impressions for each of:
• Left index
• Right index
• Left middle
• Right middle
• 4,080 samples in total
25. ZOO PLOT
• Neurotechnology Megamatcher v4.0.0
• Performix v3.1.9
• Calculated by a minutiae-based matcher
26. ZOO PLOT OVERVIEW
• Maps the
relationship
between a user’s
genuine and
imposter match
results defines four
additional classes
of worms, doves,
chameleons, and
phantoms
27. CLASSIFICATIONS OF ANIMALS
• Chameleons always appear similar to
others, receiving high match scores for
all verifications. Chameleons rarely
cause false rejects, but are likely to
cause false accepts.
• Phantoms lead to low match scores
regardless of who they are being
matched against; themselves or others.
28. CLASSIFICATIONS OF ANIMALS
• Doves are the best possible users in
biometric systems. They matching well
against themselves and poorly against
others.
• Worms are the worst users of a biometric
system. Where present, worms are the
cause of a disproportionate number of a
system’s errors.
29. ADVANTAGES OF ZOO PLOTS
OVER ROC/DET CURVES
• Traditional methods of evaluation focus
on collective error statistics such as
Equal Error Rates (EERs) and Receiving
Operating Characteristic (ROC) curves.
• These statistics are useful for evaluating
systems globally, but ignore problems
associated with individuals and
subgroups of the population. The
biometric menagerie is a formal
approach to user-centric analysis.
30. ADVANTAGES OF ZOO PLOTS
OVER ROC/DET CURVES
• In many real world situations it has been
observed that user groups performance
varies based on any number of
demographic factors.
• Researchers and system integrators are
interested in identifying which of these
groups are performing poorly as they
may be causing a disproportionate
number of verification errors.
38. FUTURE WORK
• Determine if these results hold true for
other fingerprint sensors
• Deeper analysis of the impact of poor
performing animals
39. CONTACT INFORMATION
• Michael Brockly
• Undergraduate Researcher at BSPA Lab
• mbrockly@purdue.edu
• Stephen Elliott PhD
• Associate Professor at BSPA Lab
• elliott@purdue.edu
Dove has highest count for right index and right middle. Refer back to right index and right middle having best quality and left index and left middle having best minutiae. Very small amount of RI in worms but a large amount of chameleons.
Dove has highest count for right index and right middle. Refer back to right index and right middle having best quality and left index and left middle having best minutiae. Very small amount of RI in worms but a large amount of chameleons. High chameleons can cause high false accepts.
Dove has highest count for right index and right middle. Refer back to right index and right middle having best quality and left index and left middle having best minutiae. Very small amount of RI in worms but a large amount of chameleons. High chameleons can cause high false accepts.
Dove has highest count for right index and right middle. Refer back to right index and right middle having best quality and left index and left middle having best minutiae. Very small amount of RI in worms but a large amount of chameleons. High chameleons can cause high false accepts.
Dove has highest count for right index and right middle. Refer back to right index and right middle having best quality and left index and left middle having best minutiae. Very small amount of RI in worms but a large amount of chameleons. High chameleons can cause high false accepts.