The purpose of MINEX04 was to determine the feasibility of using minutiae data (rather than image data) as the interchange medium for fingerprint information between different fingerprint matching systems. MINEX is designed to evaluate whether various populations and combinations of encoding schemes, probe templates, gallery templates, and fingerprint matchers will produce successful matches.
The MINEX04 was used to determine the feasibility of using minutiae data (rather than image data) as the interchange medium for fingerprint information between different fingerprint matching systems. The test was performed by NIST staff and was being co-sponsored by the Department of Justice (DOJ) Justice Management Division (JMD) and the Department of Homeland Security (DHS) US-VISIT program.
It has been generally acknowledged that the interchange of fingerprint image data provides the greatest interoperability between dissimilar fingerprint recognition systems. However, standards also exist that specify the location and formatting of processed minutiae characteristic data for matching purposes. There is only limited information regarding the interoperability, performance, and matching accuracy of fingerprint matching systems that use exchanged minutiae obtained from dissimilar systems. A major objective of this test was to quantify the verification accuracy changes when minutiae from dissimilar systems (rather than scanned images) are interchanged and used for matching fingerprints.
A simple model of a fingerprint minutiae recognition system may be viewed as an unknown probe template that must be compared to a known gallery template using a specific matcher to determine the similarity between the two templates. Interoperability of templates is affected by the method used to encode minutiae. There are different schemes for defining the method of locating, extracting, and formatting the minutiae information from a fingerprint image. These encoding schemes include system proprietary approaches, the FBI-IAFIS encoding approach that incorporates the number of ridge-crossings to the eight nearest neighbours, and the method specified by the ANSI M1 biometrics committee. The second factor that affects interoperability is the systems that perform the encoding for both the probe and gallery templates. Finally, the matcher used to compare the templates will have an influence on performance.
This test will determine the probability of a successful match when various combinations of the templates schemes, encoding systems, and matching systems are used. Probe templates generated by one system in accordance with one scheme of encoding, will be compared to a second template generated by a second system in accordance with the same or a different encoding scheme, using a specific fingerprint matcher.
This test is not designed to rank vendors but to determine whether various populations and combinations of encoding schemes, probe templates, gallery templates, and fingerprint matchers will produce successful matches. Each system participating was required to produce minutiae templates in accordance with their proprietary encoding scheme. This will provide a base performance level. Each vendor was also be required to produce a basic M1 template. Generation of an FBI-IAFIS template was strongly encouraged.
An amended version of the MINEX Evaluation Report is online (22 March 2006). This includes material incorrectly omitted from the initial March 6 report, see page 4.
The following supplemental documents accompany the main report:
A set of briefing slides are also available: MINEX Summary Presentation
15 organizations registered to participate in MINEX: