The performance of Automated Fingerprint Identification System (AFIS) heavily relies on how efficiently minutiae are extracted. Most, if not all, AFIS compare minutiae information (such as ridge endings and bifurcation position) in form of sets of coordinates for verification or identification. Surprisingly, research on alternative minutiae extraction schemes is scarce. This paper, proposes the implementation of the novel approach to fingerprint recognition based on the extraction of minutiae in form of circular strings, which are suitable for approximate circular string matching. In addition to that, the proposed solution is able to detect the exact location and rotation of the input fingerprint regardless of its location on the scan surface.
|Title of host publication||Patterns 2016|
|Publication status||Published - 20 Mar 2016|