Orientation Field Based Rapid Global Segmentation of Fingerprint Image

Fingerprint orientation field describes the essential texture feature of a fingerprint image,including shape,structure and direction.Fingerprint orientation field behaves similar in local region while very different in global region.We propose to segment a fingerprint image rapidly and coarsely using orientation field image,present an example of its application on local matching.We conduct experiments on FVC2004 DB1.The experimental results show that the proposed segmentation method helps to improve the identification accuracy and the efficiency of local-matching algorithms.