Abstract

Abstract. One of the long-standing problems of practice is the personal identica-tion, namely ascertaining personal identication. Several methods have been adapted for solving this problem during history, from seal ring, through signature to photo. Fingerprint, as a possibility for identication is widely known from the last century. From the sixties when computers became widely used, the possibility of computeriz-ing of identication was brought up. In this paper we explain a ngerprint identifying system used in automatic per-sonal identifying systems. The identication is based on dermatological characteris-tics, namely the unique pattern of ridges of ngers. Usually two levels of pattern are distinguished, global and local patterns In this system we improved both the global and the local characteristics using a identi cation. We use the direction matrix representing orientation of ridges as global characteristics and the method of identication of bifurcations as local characteristics. AMS Subject Classication. 68U10 Image Processing 1.