On improving interoperability of fingerprint recognition using resolution compensation based on sensor evaluation

Abstract

The purpose of this paper is the development of a compensation algorithm by which the interoperability of fingerprint recognition can be improved among various different fingerprint sensors. In order to compensate for the different characteristics of fingerprint sensors, an initial evaluation of the sensors using both the ink-stamped method and the flat artificial finger pattern method was undertaken. Then the resulted image resolution was incorporated to the compensation algorithms. This paper proposes Common resolution method and Relative resolution method for compensating different resolutions of fingerprint images captured by disparate sensors. Both methods can be applied to image-level and minutia-level. This paper shows the results of the minutiaelevel compensation. The Minutiae format adhered to the standard format established by ISO/IEC JTC1/SC37. In order to compensate the direction of minutiae in minutia-level, Unit vector method is proposed. The fingerprint database used in the performance evaluation is part of KFRIA-DB (Korea Fingerprint Recognition Interoperability Alliance Database) collected by the authors and supported by KFRIA. Before compensation, the average EER was 8.62% and improved to 5.37% by the relative resolution compensation and to 6.37% by the common resolution compensation. This paper will make a significant contribution to interoperability in the system integration using different sensors. © Springer-Verlag Berlin Heidelberg 2007.