fingerprint recognition method is described. Fingerprint recognition method according to the invention, respectively from a predetermined number of bits to produce a magnetic space (self-space) formed by a set of binary string based on the distribution and the direction component of the feature point of the input fingerprint image, the resulting magnetic space each constituting a plurality of generating MHC MHC recognition unit consisting of the detector, the method comprising: matching template fingerprint and the subject area of the image stored in the MHC detectors constituting the produced MHC fingerprint database, created with MHC recognition The detector and the method comprising: a fingerprint database of MHC binary string is matched with respect to the magnetic area of the template fingerprint image stored in the feature point to generate a local structure in the space, each represented by a center point, and the magnetic area and the matching template fingerprint image MHC and a step of determining whether or not matching of the parameters on the basis of the local structure of the detector resulting from a binary string entered by performing the matching according to the geometry of the feature point image and the fingerprint template fingerprint image. According to the present invention, by, based on the geometric structure of the magnetic recognition model and a feature point of the Artificial Immune perform fingerprint matching, it is possible to faster and more reliable fingerprint.
展开▼