Fingerprint recognition with identical twin fingerprints

Area of Science:

• Biometrics
• Forensic Science
• Genetics

Background:

• Identical twin fingerprint recognition is challenging due to genetic similarity.
• Previous studies have explored twin fingerprint similarity with limited scope.

Purpose of the Study:

• To investigate the similarity of identical twin fingerprints using advanced methods and a larger dataset.
• To analyze the probability distribution of fingerprint types among identical twins.

Main Methods:

• Utilized a database of 83 identical twin pairs (3984 images).
• Employed two state-of-the-art fingerprint identification systems: P071 and VeriFinger 6.1.
• Captured six impressions per finger for realistic score distribution analysis.

Main Results:

• State-of-the-art systems can distinguish identical twins with minimal performance impact.
• Identical twins share the same fingerprint type with a 0.7440 probability, versus 0.3215 for non-twins.
• Fingerprint type distribution and specific finger probabilities for same-type prints are consistent within identical twins.

Conclusions:

• Automatic fingerprint verification systems are effective even for closely related individuals.
• The study provides novel statistical insights into identical twin fingerprint characteristics.
• Findings contribute to understanding biometric uniqueness and twin-specific patterns.