研究目的
Investigating the phenomenon of template aging in iris recognition and its impact on recognition error rates.
研究成果
The study confirms the existence of iris template aging, showing an increase in FNMR with time lapse. It suggests that controlling pupil dilation and updating enrollment templates can mitigate this effect. The findings challenge the notion that a single iris enrollment can last a lifetime.
研究不足
The study's findings are based on a specific sensor and matcher, and the generalizability to other systems may vary. The impact of other aging-related factors beyond pupil dilation was not fully explored.
1：Experimental Design and Method Selection
The study evaluates the impact of time lapse on iris recognition accuracy using a dataset acquired over three years with an LG 4000 sensor. The VeriEye matcher was selected for its superior performance among four evaluated matchers.
2：Sample Selection and Data Sources
Iris images were acquired from 2008 through 2011, involving 322 subjects with varying demographics. The dataset includes images from different years to analyze short-term and long-term matches.
3：List of Experimental Equipment and Materials
LG 4000 iris sensor, VeriEye SDK (version 2.4) for iris matching.
4：Experimental Procedures and Operational Workflow
Images were acquired over multiple years, with matches categorized into short-time-lapse (within a few months) and long-time-lapse (different years) comparisons. The FNMR was analyzed across different time lapses.
5：Data Analysis Methods
The FNMR was analyzed as a function of the decision threshold, with bootstrap methods used to estimate confidence intervals for the change in FNMR.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
