Characterizing habituation using the time-on-task metric in an iris recognition system
Abstract
This thesis presents a characterization of biometric habituation in an iris recognition study using qualitative analysis of a distributed habituation survey and quantitative analysis of iris images collected in 2010 and 2012. The performed analyses answered the following two questions: a) How consistently does the biometric community define habituation?; and b) Does the time-on-task variable provide enough evidence to indicate the existence of habituation in an iris recognition system? The qualitative analysis examined responses to 12 habituation-related questions from 13 biometric experts to identify common themes that not only determined definition consistency but also characterized critical components often omitted from habituation definitions. Upon completion of the survey analysis, this study concluded that while aspects of habituation were universally understood, habituation in its entirety was not. The quantitative analysis examined trends in mean time-on-task using number of visits as a covariate. Subjects repeatedly (20 captures per visit and 25 maximum attempts per visit) interacted with an iris recognition camera, returning for at least eight visits. The trends in the resulting time-on-task, image quality and matching performance indicated that habituation effects were identifiable near the end of the 2012 collection.
Degree
M.S.
Advisors
Elliott, Purdue University.
Subject Area
Biomedical engineering|Systems science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.