Paper Authors

Abstract

NOTE: The first page of text has been automatically extracted and included below in lieu of an abstract

Laboratory Exercises for an Undergraduate Biometric Signal Processing Course

Introduction

The ability of investigators to analyze sound, image, and video data and to efficiently search through large databases of biometric data such as fingerprints or facial images has revolutionized the field of forensics over the last couple of decades. These technologies are routinely used in popular television shows such as Crime Scene Investigation and NCIS among others. Indeed, it is seldom that one of these shows does not feature a biometric identification such as fingerprint identification, face identification, or speaker identification.

Since September 11, 2001, there has been a much greater emphasis placed worldwide on security, and biometric technologies have been adopted to enhance existing security measures. Biometrics such as fingerprints and a digital image are stored in an RFID chip embedded in the passports of citizens in a large number of European countries. The United States has introduced the US-VISIT program and gathers face and fingerprint data from travelers entering its borders1. While governments have acted to increase security, access to both public and private buildings and sensitive areas are being more rigorously controlled and biometric technology is being more widely used for identity verification. As one example, the successful use of fingerprinting at Walt Disney World shows that a biometric system can be used in a large commercial setting and is readily accepted by people2. Biometric technology has also been incorporated into popular laptops from Lenovo3 among others.

Although biometric technology is increasingly important in forensics and security applications and is currently a very active research area, this topic has predominantly been addressed at the graduate level due to the mathematical background required for research in the area. In a recently funded NSF proposal4, we have proposed the development of a biometric signal processing course targeted towards undergraduate students. An important component of the project was the proposed development of a set of laboratories which would give undergraduate students in Electrical and Computer Engineering (ECE) the opportunity to explore these important technologies. In this paper, we detail these newly developed laboratories (based on a 10 week term) that allow students to gain hands-on experience with real-world biometric technologies.

Each set of laboratories is based on a different clue and the ultimate goal of each laboratory experiment is to identify the persons responsible for an imaginary crime. The first three laboratories are based on a speech signal clue and this gives students an opportunity to review material from their introductory signal processing course while also introducing the building blocks of a pattern recognition system: signal preprocessing and enhancement, feature extraction, and classification. The second set of three laboratories focuses on face recognition. The three stages of the recognition task are reinforced by breaking the task into separate experiments, and in each laboratory one part of the recognition system is implemented. In the third set of laboratories, a fingerprint image is presented as a clue and students develop some of the code for the fingerprint recognition system. In the final couple of weeks, student groups are assigned different projects that modify the basic classifiers developed during the previous weeks. Students implement these projects and determine the effect of the project on recognition performance.

• Citation

• Format
  • APA
  • APA - LaTeX bibitem
  • MLA
  • MLA - LaTeX bibitem
  • Bibtex
  • EndNote - RIS

Cotter, S. (2010, June), Laboratory Exercises For An Undergraduate Biometric Signal Processing Course Paper presented at 2010 Annual Conference & Exposition, Louisville, Kentucky. 10.18260/1-2--16478