방향정보와 특징점을 사용한 지문영상의 형상 분류에 관한 연구

A Study on Fingerprint Classification Using Directional Information and Singular Points

In this paper, a fingerprint classification algorithm is presented. Fingerprint types are classified into five categories: arch, tented arch, left loop, right loop and whorl. Singular points (cores and deltas) are detected using Poincare index on the directional image smoothed by adaptive window size. The method is shown to be robust to the variation of fingerprint image qualaity.