This thesis is about the research on extraction minutiae in AFIS(automatic fingerprint identification system) and AFAS(automatic fingerprint authentication system) which get into the spotlight these days in the area of user identification and authentication.
The study of AFIS and AFAS has deeply investigated since 1960. But the performance is not guaranteed when the digital fingerprint image is not in good quality.
So our objective in this thesis is to design new minutiae extraction method which is more accurate and more confident independent of the quality of fingerprint image.
In this Thesis, we have proposed and analyzed the followings : (i) fuzzification of fingerprint image, (ii) new minutiae extraction method on fuzzified fingerprint image, (iii) new filtering method of directional value in fingerprint, (iv) ridge following method.
Our system have been evaluated compared with classic minutiae extraction method and original ridge following method. Ten inked fingerprints and ten NIST DB4 fingerprints are used in the evaluation. And experimental results show that our systems perform better on these data.
