Reproducibility and repeatability experiment with nested factors applied to the evaluation of a fingerprint...

Area of Science:

• Forensic Science
• Measurement Science
• Biometrics

Background:

• Accurate fingerprint morphometric measurements are crucial for forensic identification.
• Traditional gage R&R methods may not fully capture complex error sources in biometric data.
• Understanding variation components is key to improving fingerprint analysis reliability.

Purpose of the Study:

• To introduce a redefined gage reproducibility and repeatability (R&R) method for fingerprint morphometric systems.
• To identify and quantify error sources and variation components in ridge width measurements.
• To leverage linear mixed effects modeling for a more robust R&R analysis.

Main Methods:

• Development of a redefined gage R&R method incorporating linear mixed effects models.
• Analysis of ridge width data from four types of finger impressions (inked-rolled, inked-flat, latent white powder, latent black powder).
• Data collection involving 10 donors, four examiners, and multiple repeated measurements per impression (24-36 times).

Main Results:

• The redefined method effectively decomposes variation sources in fingerprint ridge width measurement.
• High repeatability (low intra-examiner variation) was observed across examiners.
• Low reproducibility (high inter-examiner variation) was identified, indicating potential for improvement in consistent measurement application.
• Minimal skin distortion was found to impact ridge width measurements.

Conclusions:

• The proposed linear mixed effects-based gage R&R method provides a powerful framework for analyzing measurement system variability.
• Inter-examiner variability is a significant factor affecting the reliability of fingerprint morphometric measurements.
• The variance decomposition approach is effective for pinpointing specific sources of error in biometric data acquisition.