(Fall 2011) IT 345 Posters
Upcoming SlideShare
Loading in...5
×
 

(Fall 2011) IT 345 Posters

on

  • 203 views

 

Statistics

Views

Total Views
203
Views on SlideShare
189
Embed Views
14

Actions

Likes
0
Downloads
0
Comments
0

1 Embed 14

http://www.bspalabs.org 14

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

(Fall 2011) IT 345 Posters (Fall 2011) IT 345 Posters Document Transcript

  • How Hard is a Signature to Forge?Joseph O’Neill, Dr. Stephen Elliott, Dr. Richard Guest, , Kevin O’ConnorThe purpose of this study is to take the idea of forging signatures from one that is centered on the experience ofthe forger, and looking at the characteristics of the signature itself. This was done by collecting data in the form ofmultiple surveys from a selected group of semi-experts in the field of biometrics. These surveys were created tofocus on specific aspects of the signature that forgers would use to replicate the signature. The measures werethen correlated to determine the difficulty level of the survey. Many of the other features were studied to finddifferent results as well. The underlying thesis is before generating an impostor distribution from forgers, youshould examine the forgers perception of difficulty.1. Next stage is examinecorrelation between theperceived attributes and theextracted features from thesignature.2PERFOverview3PERFPhase 2Initial Survey QuestionInitial steps were to assess the opinion of a set of non-professional forgers on a signature.After initial analysis, the metrics were further refined andan additional survey was created, and tested on a LikertScaler024681012141618Signature 111 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)024681012141618Signature 112 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0510152025Signature 114 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0510152025Signature 116 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)02468101214Signature 104 (C)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)02468101214Signature 119 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 122 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 126 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)02468101214Signature 128 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 129 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)024681012141618Signature 132 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 136 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 140 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)024681012141618Signature 152 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 143 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)02468101214161820Signature 173 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 183 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)0246810121416Signature 184 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)02468101214161820Signature 190 (E)Simple - ComplexIllegible - LegibleSloppy - NeatStraight - CurvedCommon - Unique(a) - (b)Initial Results from the first survey
  • Impact of Age and Gender on Fingerprint RecognitionSystemsLindsay Sokol, Carson Weaver, Cassandra Harrell, Roy Mills, Brandon Galbreth, Steve McKinney, KevinO’Connor, Stephen ElliottThe purpose of this study is to investigate the impact of subject age and gender on fingerprint performance,quality, and image characteristics. Previous research has shown that both age and gender impact fingerprintimage quality and performance. By identifying problem populations, recommendations can be made to improvethe interaction for these populations, in turn improving the performance of the entire system. This study willuse previously collected data using a ten-print, live-scan fingerprint sensor.Previous Research Image QualityPreliminary Results1. Examine performance if gender and age at variouslevels2. Examine the role of entropy and gender5Overview6Phase 2Data Cleaning ChallengesPaper Ref. Performance HenryClassificationImage Quality Minutiae Ridge Width Entropy[2] DifferenceF better thanMDifferenceM betterthan FNodifference[3] Nodifference[4] Difference[5] SlightDifference[6] SlightDifferenceF better thanMSlightDifferenceDifferent,except for RIM betterthan FNoDifferenceexcept forRRThisresearch Phase 2  Phase 1  Phase 1  Phase 2Image Quality variables are broken down into differentcategories:• Good – represents the part of the image whereminutiae points can be reliably extracted• Poor – broken and smudged pixels• Light – a count of light areas• Dark – a count of dark areas• Minutiae – the number of minutiae found in the centralarea of the fingerprintTypically, we have not analyzed image quality as part ofour performance testing previously.Design of Experiments classexamined the dataset as partof the course.The datasets were skewed –thumbs contributed to someof the image quality issues –especially in Light, Dark,Good, Poor.The group had to understanddistribution and had to makesuggestions about the data.Understanding and cleaningthe data took the majority oftime in Phase 1.Hypothesis Performance Image Quality Minutiae Core DeltaNo difference incore/deltaacross genderP=0.000No difference incore/deltaacross age(h)P=0.000No difference incore/deltaacross age(m)P=0.000No difference inminutiae acrossgenderP=0.365No difference inimage qualityacross genderP=0.000No difference inimage qualityacross age(h)P=0.000No difference inimage qualityacross age(m)P=0.000Withrespect togenderPerformance HenryClassificationImage Quality Minutiae Ridge Width Entropy Core DeltaThisresearch Phase 2 Difference(global) Phase 2examinefingerlocation.Nodifference Phase 2 Difference
  • Design and TestingIdeationOur original concept was shamelessly stolen from themovie “Men In Black”, where fingerprint verification isdone using a ball-shaped scanner, scanning all fingers atthe same time. Studies have shown user comfortimproves scan quality, so we decided to adapt theexisting fingerprint scannerto improve ergonomics,increase throughput time,and see an improvementin image quality as anadditional benefit.Measurement & FrameworkOverviewOptimizing Interaction Time For Fingerprint VerificationThomas Cimino Brandon Hilts Chris Clouser1Biometric Standards, Performance, and Assurance Laboratory, Department of Industrial Technology2Department of Psychological Sciences, Purdue UniversityVariables Of Interest10-Print Fingerprint ScannerHuman Interaction&Observation AreaAnalysisUsability AnalysisStatistical Data ComparingImage QualitySatisfaction•QuestionnaireoUser FeedbackEfficiency• TaskTimeEffectiveness• Number of ErrorsFingerprint Image Quality Analysis Biometric System Performance• The initial goals of this project were to decrease the amount of time a user spends going through a fingerprint scan for normal uses (border control,facility access, etc.) without losing any of the image quality. It is our hypothesis that we can reduce overall user interaction times by approximately one-third by adopting this design, as it will eliminate the user needing to reposition the hands once they begin the scanning process.EAF-3AD
  • Ergonomic Improvements to Hand Geometry ReadersRoss Barbish, Chuck Oliver, Rob Larsen, Narut Chitrudi-Amphai, Markus Jones, Tera Engle, KevinO’Connor, Stephen ElliottPrototype #1 Prototype #2 Prototype #3The purpose of this study was to weigh the objective performance decrease with the subjective comfort increasewhen ergonomic accessories were attached to the surface of the hand geometry readers. These accessories arecorrectly assumed to have a negative effect on performance, but the question is whether or not the degree towhich performance is decreased is acceptable or not. Our research has found that the performance decrease issmall enough that these accessories are viable options to improve the ergonomics of these devices. There alsois a near consensus among users as to which accessories are more comfortable. This is exciting as futureresearch and development can focus on the material and comfort of these accessories.Identification of VariablesInitial Phase 1 results (non MSD)Devices and PrototypesDesign of Experiment1. Next stage is to test the three different prototypes in ascenario and operational environment2. Prototypes will be deployed on the door of Knoy 378and in a test cell in MGL testing lab3. MSD population will be recruited after IRB approval5Overview6Phase 2• Scores were relatively stable with and without ergonomic attachments.• Minimal fluctuations of our average scores between test runs.• Ergonomic improvements placed underneath the hand definitely a possibility.• Hand scores will be collected from a randomly selected group ofMSD and Non-MSD subjects on all 3 prototypes and withoutprototype• Readings on each will be collected for each participant.• A comfort index score will be collected for each reading.Run ChartComfort index – notused in phase 1, but willbe used in phase 2
  • Understanding Environmental ConditionsCharles Belville, Jason Wintz, Andrew Thomas, Stephen Elliott, Mitch MershonThe purpose of this study is to measure the environmental conditions in a testing lab, and to provide guidance toISO /IEC JTC 1SC 37 working Group 5. Biometric technologies are impacted by environmental conditions – forexample face recognition and lighting. However, no methodology exists to measure environmental conditions in abiometric testing lab. The output of this project will be to contribute documents and test methodologies to SC 37as well as implement environmental monitoring for data collection in the Spring.1. Next stage is to replicate the study in Knoy 3782. Provide guidance to the biometric community on how to setup an environmental study3. Contributions will include input to ISO/IEC JTC 1 SC374. Continue with the teleconferences with the Spanish editorial team8PERFOverview9PERFPhase 2Preliminary Test DesignMGL B307 – 30 (4’X4’) ZonesTest Period in each Zone: 2 HoursSampling Rate: 120 seconds.There will be a total of three different tests performed in MGLB307.The first test will record:• illumination,• temperature,• humidity,• pressureThe second test will record sound at rest occupancy stateThe third test will record sound at operational occupancystate.Test Plan1. Setup EN300 data logging to SD card2. Setup SD700 data logging to PC via RS2323. Begin data collection in Zone 11. Ensure lights are stable (approx: 6 minutes)2. Ensure room is empty of all personnel4. Move data collection tripod to each zone every twohours.5. Compile data from each zone once completeOther Zone ResultsZone Results