A non-linear model for fingerprints matching based on minutiae-core pairs interaction

In this paper, we propose a non-linear model for fingerprints matching which handles more efficiently the nonlinear deformation problem in fingerprint matching. It is based on the fact that a fingerprint image is composed of different minutiae with various orientations separated by a spatial distance that can be successfully represented by an interaction vector. After extracting the minutiae and the fingerprint core point we determine the orientation of the ridges around the core point and each minutia. The distance and orientation between the core point and each minutia are calculated and used in the interaction model to obtain the minutiae-core interaction index. A distance metric is used to compare the interaction descriptors of two fingerprints. The results of the comparisons are used to calculate the minutiae-core index of the fingerprint. Our non-linear model is experimented on FVC2000 DB1 and we feel it is promising although the results are fairly good.