A variational formulation for fingerprint orientation modeling

摘要: Orientation pattern is an important feature for characterizing fingerprint and plays critical roles in recognition classification. This paper proposes a framework modeling the orientation field based on variational principle, where can be estimated through solving associated Euler-Lagrange equation. Compared with existing methods, our proposed method has following features. Firstly, it does not require any prior information about structure of acquired fingerprint, such as location singular point(s). Secondly, explicitly provides freedom singularity field. Thirdly, less number parameters. Comparison been made respect to state-of-the-arts terms accuracy, enhancement point detection. Advantages are demonstrated.