Premium
Performance Evaluation of Automated Fingerprint Identification Systems for Specific Conditions Observed in Casework Using Simulated Fingermarks
Author(s) -
de Jongh Arent,
Rodriguez Crystal M.
Publication year - 2012
Publication title -
journal of forensic sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.715
H-Index - 96
eISSN - 1556-4029
pISSN - 0022-1198
DOI - 10.1111/j.1556-4029.2012.02114.x
Subject(s) - minutiae , computer science , fingerprint (computing) , identification (biology) , artificial intelligence , pattern recognition (psychology) , distortion (music) , relevance (law) , fingerprint recognition , biology , amplifier , computer network , botany , bandwidth (computing) , political science , law
Few studies have been reported on the performance evaluation of automated fingerprint identification systems (AFIS) for fingermark‐to‐fingerprint comparisons. This paper aims to illustrate to fingerprint examiners the relevance of evaluating the AFIS performance under specific conditions by carrying out five types of performance tests. The conditions addressed are the number of minutiae assigned to a fingermark, manual and automatic assignment of the minutiae, the finger region from which the fingermark originates, the degree of distortion in the fingermark, and the difference in orientation between fingermarks and fingerprints. In these tests, the magnitude of the influence for each condition was quantified. The comparisons were performed using a research AFIS technology with simulated fingermarks. Simulated fingermarks provide a practical way to create fingermarks for specific conditions in large quantities. The results showed that each condition influences the performance significantly, emphasizing the relevance of developing, and applying performance tests for specific conditions.