Animal Biometrics

Today we post a new article on "An Evaluation of Retinal Imaging Technology for 4-H Beef and Sheep Identification". Please add your comments.

Research Ideas

We are always interested in new research ideas. Have one? Please leave your suggestions here

Comments on our website

We would like your feedback on our website. If you have any comments, please let us know.

Equating Biometric Entropy

This research is being conducted as a part of a Masters thesis by Matthew Young; advised by Stephen Elliott, Ph.D. For more information, please visit our website: http://www.biotown.purdue.edu

The main objective of this study is to determine the key space and entropy in bit strength of fingerprint biometric technologies by assessing probabilities of minutiae locations in a fingerprint image.

Abstract: The use of biometric characteristics for access control to both physical and logical resources has scene tremendous growth in recent years. The increased potential of biometric authentication has been accelerated by the inherent qualities of tightly binding the authenticator to the identity of an individual user. Uniqueness of biometric characteristics varies depending upon which biometrics modality is being evaluated, but one common theme is true and that is the authenticator is a digital representation of the characteristic provided by the individual. This factor makes it virtually impossible to determine how many real life possible representations of biometric data could be presented to the systems. In order to accomplish that task, every human being in the world would have to be considered in such a process.

Previous work in this area as shown by Ratha, Connell, and Bolle, has focused on defining keyspace for biometrics; specifically fingerprints. Keyspace being the total number of possible values of keys in a cryptographic algorithm or other security measure such as a password. Applying this definition to biometrics, key space is the total number of possible features, minutiae points for fingerprints in this case, that could exist in a biometric sample. The previous approach segmented a fingerprint image into possible locations for minutiae to estimate the total possible values, giving equal probability weighting to all locations [1].

Methodology: This research looks to build on the work shown above by determining probability of occurrence for individual potential minutiae locations. Just as some passwords and secrets contain values that are more likely to occur than others, so too are features in biometrics samples [2]. Referring to fingerprints in this case, it is visually evident that a greater number of minutiae are likely to be around the center of the rectangular image rather than unoccupied corners.

In order to equate entropy of biometrics similarly to passwords based on probability of potential values; the principles of Shannon’s Information Theory will be applied to minutiae in fingerprints. Shannon summarizes entropy as the randomness of the information contained in a message [3]. Shannon’s equation for determining entropy as a value of H in bits is shown below:

Apply this to fingerprint minutiae:
• n = total # of possible locations for minutiae in the image.
• p(X) = the probability of minutiae occurring at each individual location.

To accomplish the task of calculating probabilities for minutiae locations:
3 fingerprints on 8 fingers from 255 subjects will be examined using commercially available fingerprint software.
Adjacent segments, determined by the common area minutiae points consume [1], of dimension (15x15 pixels) will be aligned within the standard image dimensions of (248x292 pixels) for the dataset being examined.
Minutiae points for will be recorded as present or not present in the appropriate segments based on their location of coordinates (x,y).
After processing all images in the data set, percentages for actual occurrence of minutiae in all possible segments can calculated.
Based on the these percentages, The required parameters of Shannon’s entropy equation are now available and can be used to equate the entropy of fingerprints, represented in bit strength.

Blog Post References:
1. Ratha, N., J. Connell, and R. Bolle, Enhancing security and privacy in biometrics-based authentication systems. IBM Systems Journal, 2001. 40(3).
2. Doddington, G., et al. Sheep, Goats, Lambs and Wolves. An Analysis of Individual Differences in Speaker Recognition Performance. in International Conference on Spoken Language Processing. 1998. Sydney, Australia.
3. Shannon, C.E., Communication Theory of Secrecy Systems. Bell Systems Technical Journal, 1949. 28: p. 656-715.

Carnahan Conference 2006 - An Assessment of Dynamic Signature Forgery and Perception of Signature Strength

Please post any comments, questions, and suggestions

Carnahan Conference 2006 - Keystroke Dynamics Verification using a Spontaneously Generated Password

Please post comments below on this paper presentated at the Carnahan Conference

Carnahan Conference 2006 - Perceptions of Retinal Imaging Technology for Identifying Livestock Exhibits

Please post your comments on the Carnahan Conference presentation and paper

Carnahan Conference 2006 - Implementing Ergonomic Principles in a Biometric System: A look at the Human Biometric Sensor Interaction

Abstract: This paper discusses the implementation of ergonomic principles in a biometric system. Historically, the biometrics community has performed limited work in the area of ergonomics and usability. This research discusses an experiment involving a swipe fingerprint sensor which examined the human interaction with the biometric device called the Human Biometric Sensor Interaction (HBSI). The purpose of this study was to examine issues related to fingerprint acquisition of all ten digits. The results revealed that there are fingerprints that have higher Failure to Acquire (FTA) rates than others, which reveals that more research is needed in the area of biometric usability and ergonomics, namely understanding how the human interacts with the biometric sensor.

Please post any comments, questions, or suggestions.

Changes in Finger Preferences Over a 6 Week Period of Interaction

Abstract: The Human Biometric Systems Interaction (HBSI) is a field that focuses on user perceptions and preferences. This project is to establish whether the user is likely to change their preferred fingers after interacting with an optical fingerprint sensor over a time frame of 6 weeks.

Methodology: The methodology involves having subjects interact with a commercially available 500 dpi optical sensor to acquire 10 images from their three preferred fingers during week 1 of the experiment and then 3 images during subsequent interactions for 6 weeks. After the final image acquisition, the users will be asked to re-rank their preferred fingers. In order to evaluate the changes in preference from pre to post interaction, a series of Chi-Squared tests will be run on the three finger preference rankings.

Add your comments to the methodology, and discuss this research project

Image Quality and Temperature Before and After Repeated Fingerprint Samples

Abstract: This project is to establish whether the temperature of the finger changes over repeated samples, and whether there are any ramifications to image quality and fingerprint matching performance. The analysis statistically analyzed the image quality between the two (before and after images)

Methodology: The methodology involves having subjects interact with a commercially available 500 dpi optical sensor to acquire three images of the same dominant index finger which will also be used to record temperature. The user would then interact with eight other sensors (seven of them swipe, and one optical) using their dominant index finger on all sensors. The last sensor that the individual will interact with will be an optical sensor of the same model used at the beginning of the experiment. Similar to before, the subject will provide three additional images. After the final sensor, the temperature of the dominant index finger will be recorded again.

A commercially available image quality software will be used to calculate image quality for the fingerprint samples. In order to test if the image quality is the same before and after interaction, Wilcoxon Signed Rank Test for Paired Comparisons, which is a non-parametric test will be performed. VeriFinger 5.0 by Neurotechnologija will be used to measure False Non Match Rate (FNMR) and False Match Rate(FMR).

Add your comments to the methodology, and discuss this research project

Performance Consistency of Fingerprints Across Different Sensors

Abstract: Large scale fingerprint recognitions implementations will move towards an architecture with centralized matching/storing subsystem and a decentralized accquisition subsystem. In such an implementation it is necessary to understand consistency issues of fingerprints collected from the same individual across different fingerprint sensors. This study is aimed at analysing consistency of image quality and matching performance rates for fingerprint samples collected from three different sensors.

Methodology: Three different fingerprint sensor were used to perform data collection
DigitalPerson U.are.U 4000 -optical sensor 500dpi
Identix DFR2090 -optical sensor 500dpi
Authentec AF-S2 -capacitance sensor 250dpi

6 fingerprint images from the right index finger were collected from 55 subjects on each of the three sensors.

A commercially available image quality software will be used to calculate image quality for the fingerprint samples. In order to test if the image quality is the same across the three different sensors, Kruskal -Wallis test will be performed.
VeriFinger 5.0 by Neurotechnologija will be used to measure False Non Match Rate (FNMR) and False Match Rate(FMR).

Add your comments to the methodology, and discuss this research project

Image Quality and Performance of Fingerprints Across Age Groups

Abstract: Fingerprint recognition systems are heavily influenced by the quality of samples provided to the system. Various factors like ability to interact with the sensor, wear and tear on surface of the finger, elasticity of the skin on finger etc. influence the quality of fingerprint samples. With deployments of fingerprint recognition systems that include users from all types of age groups an understanding of age on fingerprint image quality and performance rates is imperative. The focus of this research is to study the difference in image quality and performance rates across four different age groups: 18-25, 26-39, 40-62, 62 and above.

Methodology: DigitalPersona U.are.U 4000 optical sensor was used to collect the fingerprint images. Three fingerprint samples were collected from each finger in the dataset. There were unequal number of different fingers in each age group. Given below is the number of different fingers in each age group:
18-25: 158 different fingers
26-39: 48 different fingers
40-62: 52 different fingers
62 and above: 120 different fingers

A commercially available image quality algorithm was used to extract image quality score and minutiae count for the fingerprint samples.

The analysis will invovle performing non-parametric Kruskal-Wallis test for difference in image quality and minutiae count between the 4 groups. Post-hoc analysis will be performed if image quality and minutiae counts are found to be different across the age groups. The performance rates will invovled measure Fasle Non-Match Rates(FNMR) and False Match Rates(FMR).

Add your comments to the methodology, and discuss this research project

Dynamic Signature Verification forgery level

This project aims at understanding the levels of forgery that exist for dynamic signature verification. Here is what we propose:

levels 0-2 are "blind" forgeries (forger does not have access to copy of signature)
levels 3&4 are static forgeries (forger only has access to signature after it is signed, does not observe signature creation)
levels 5&6 are observation based forgeries (forger actually observes signature creation)
levels 7&8 are assited forgeries (either victim or technology assisted forgeries)

Possible new definitions of forgery levels

0 = zero effort, forger signs a random name
1 = forger has heard name but not seen it in print (does not know spelling Steven vs. Stephen or Jon vs. John)
2 = forger has seen name in print such as a phone directory or business card but has not seen signature
3 = single sample, forger has access to a single sample of victim's signature (receipt or check)
4 = multiple sample, forger has access to multiple samples of victim's signature (possibly additional sample of writing such as a hand written note)
5 = single observation, forger has observed the victim signing his or her name once
6 = multiple observation, forger has observed the victim signing his or her name multiple times (possibly video tape of signing replayed over and over)
7 = "victim assisted" forgery, victim intentionally coaches forger to dynamically immitate signature
8 = "technology assisted" fogery, forger has access to digitizer output of victim's signature and is allowed multiple practice attempts to immitate speed, pressure and curves

The Relation of the Human Biometric Sensor Interaction to Dynamic Signature Verification

Abstract: Many different types of point of sale digitizers are available on the market, each having their own particular characteristics such as size of the signing area, sampling speeds, and ergonomic design. Some digitizers provide real-time feedback in the form of electronic ink, while others provide none. The purpose of this paper is to establish whether there are any differences in the signing behaviors of an individual when the signing space for the signature is changed, and whether the features are significantly different depending on the presence or lack of real-time electronic ink feedback. This research was undertaken to provide information to the biometric community, specifically the ISO biometrics testing committee regarding modality testing of dynamic signature verification (DSV).

Methodology: Two different studies were undertaken to understand the issues related to the act of signing in different sizes of signing areas, as well as signing with or without ink feedback. For the first study, the subjects signed their names once in each of three different size signing areas on a Wacom Intuos™ digitizer. The three signing areas were 150mm x 200mm, 55mm x 55mm, and 55mm x 37mm. The features were extracted from the digitizer, and were subsequently analyzed. The second study involved a different set of individuals who signed their name on a digitizer that offered visible ink feedback (see Fig. 2, left) and on a digitizer that offered no visible ink feedback (see Fig. 2, right). Features were extracted from the digitizers and subsequently analyzed using a two-sample t-test. Table I outlines the feature variables used in the second study with the ink/no ink feedback digitizers.

Add your comments to the methodology, and discuss this research project

Image Quality and Minutiae Count Comparison for Genuine and Artificial Fingerprints

Abstract: The vulnerabilities of a biometric sensor have been discussed extensively in the literature, and popularized on many films and television programs. The focus of this research is to examine the image quality of an artificial print as compared to a genuine finger, and to examine the characteristics of the two including minutiae counts, image quality, as repeated samples are taken from the fingers. For the full article, please go to: http://www.writely.com/View?docid=dhcxfks5_32d37f2b