Minutiae extraction using improved Binarization process of the fingerprint

지문의 개선된 이진화 과정을 통한 특징점 추출

  • 손원무 (경성대학교 전기전자공학과) ;
  • 송종관 (경성대학교 전기전자공학과) ;
  • 윤병우 (경성대학교 전기전자공학과) ;
  • 이명진 (경성대학교 전기전자공학과)
  • Published : 2004.10.01

Abstract

Automatic fingerprint identification is a process of direction extraction, binarization, thinning, minutiae extraction of fingerprint identification. In this process, binarization after direction extraction affects a process of thinning and minutiae extraction. The fasle binarization is increased the false minutiae extraction rate. In this paper, we proposed more exact minutiae extraction algorithm with more enhanced binarization method, compared with traditional binarization process. We could have more enhanced results by using the direction and the half distance between ridges as the threshold of binarization process. In an experiment, Fingerprint images from NIST DBI are tested and the result shows that the proposed binarization algorithm increases minutiae extraction.