Fingerprint Image Enhancement Based on Morphological Filter

Fingerprint image enhancement is a essential preprocessing step to ensure good performance in Automatic Fingerprint Identification System (AFIS). In this paper, we proposed a new fingerprint enhancement algorithm based on morphological filter. This algorithm first employs the idea of median filtering based on the Mehtre method to obtain a reliable ridge orientation field. And then it designes a morphological structuring element database to save structuring element templates of all directions. Finally, in order to remove the empty and rupture in the ridge, the morphological closing operation is employed through selecting the corresponding template according to the local ridge orientation. We have tested the algorithm on FVC2004 fingerprint database DB2. Experimental results indicate that the proposed algorithm can obtain great results at both the orientation field estimation and the minutiae detection performance.