An application of second derivative of Gaussian filters in fingerprint image enhancement

Abstract

In this paper we demonstrate the use of second derivative of Gaussian filters (SDG) together with pyramid technique for fingerprint image enhancement. The quality of enhanced image is measured by the success in core point detection employing poincare method. DB1-A of FVC-2004 database is used as an input in our experiment. It is found that the SDG works well and can offer the similar filtering performance as another directional filter: Gabor filter. Once the SDG is employed in the pyramid technique, the enhanced image is superior to that of applying SDG only (without pyramid scheme). ©2010 IEEE.