摘要：Fingerprints have been significantly utilized for enormous security systems for their high-level accuracy. They even have been used effectively in smartphone devices for payment in addition to the security purpose. In this paper we propose a new finger print recognition method based on ridge and valley features in finger print images. These features extracted by rotating Gaussian semi _filters. The novelty of this finger print approach based on the mixing of directional filters and differences of Gaussian filter ideas. We obtain a new ridge/valley detector enabling very precise detection of ridge/valley points. This method conducted on different fingerprints database images, the proposed method achieves recognition accuracy rate of more than 99% for three fingerprint image datasets ATVS, LivDet2009, and LivDet2011.
关键词：Fingerprints;Features Extraction;Ridge And Valley Detection;Gaussian Semi _Filters
