Fusion of Face and Iris Biometrics Using a Stand-Off Video Sensor

Doctoral Dissertation
Thumbnail

Abstract

Multi-biometrics, or the fusion of more than one biometric modality, sample, sensor, or algorithm, is quickly gaining popularity as a method of improving biometric system performance and robustness. Despite the recent growth in multi-biometrics research, little investigation has been done to explore the possibility of achieving multi-modal fusion from a single sensor. This approach to multi-biometrics has numerous advantages, including the potential for increased recognition rates, while still minimizing sensor cost and acquisition times. In this work, experiments are presented which successfully combine multiple samples of face and iris biometrics obtained from a single stand-off video sensor. Several fusion techniques are explored to test the effectiveness of multi-modal and multi-algorithm fusion, with the best recognition rates achieved by using a Borda count of face and iris modalities. The final results out-perform either single-modality approach, and the proposed multi-biometric framework represents a viable and natural extension to many commerical stand-off iris sensors.

Attributes

Attribute NameValues
URN
  • etd-04172012-182323

Author Ryan Connaughton
Advisor Kevin Bowyer
Advisor Patrick Flynn
Contributor Kevin Bowyer, Committee Co-Chair
Contributor Nitesh Chawla, Committee Member
Contributor Karen Hollingsworth, Committee Member
Contributor Patrick Flynn, Committee Co-Chair
Contributor Scott Emrich, Committee Member
Degree Level Doctoral Dissertation
Degree Discipline Computer Science and Engineering
Degree Name PhD
Defense Date
  • 2011-12-09

Submission Date 2012-04-17
Country
  • United States of America

Subject
  • face recognition

  • biometrics

  • iris recognition

  • multi-biometrics

  • biometric fusion

Publisher
  • University of Notre Dame

Language
  • English

Record Visibility and Access Public
Content License
  • All rights reserved

Departments and Units

Files

Please Note: You may encounter a delay before a download begins. Large or infrequently accessed files can take several minutes to retrieve from our archival storage system.