An intelligent face features generation system from fingerprints


SAĞIROĞLU Ş., ÖZKAYA N.

TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES, cilt.17, sa.2, ss.183-203, 2009 (SCI-Expanded) identifier identifier

Özet

In this study, a novel intelligent system based on artificial neural networks was designed and introduced for generating faces from fingerprints with high accuracy. The proposed system. has a. number of modules including two feature extractions modules for acquiring the fingerprints and faces into the system, two feature extractors for extractors the feature sets of fingerprint and face biometric, an artificial neural network Module that was configured with the help of Taguchi experimental design method for establishing relationships among the biometric features,. a face re-constructor for building up face features from the results of the system, and a test module for test the results of the system. 10-fold cross validation technique was used for evaluating the performance of the system. The results have shown that the face features can be successfully generated from, only fingerprints. It can be concluded that the proposed Study significantly and directly contributes to biometrics and its new applications.