Fingerprint feature extraction.
Tin Tin Aye.
Date of Issue2003
School of Electrical and Electronic Engineering
This dissertation focuses on extraction of minutiae from the gray fingerprint images directly without going through the usual binarization and thinning processes. Multilayer feedforward neural network is used for extraction of minutiae. Two types of most commonly used minutiae, bifurcations and ridge endings, are extracted. If the feature is not bifurcation or ridge ending, we name this image as the 'none' image class.
DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Biometrics
Nanyang Technological University