Classifying fingerprint images using neural network: deriving the classification state

A neural network is constructed for classifying fingerprint images. The two-step learning method is proposed as a learning process, together with the four-layered neural network which has one subnetwork for each category. The classification results for 500 unknown samples are 86.0% classification rate for the first candidate and 99.0% classification rate including the second candidate. The principle component analysis is carried out with respect to the unit values of the second hidden layer, and the fingerprint classification state represented by the internal state of the network is studied. It is confirmed that the fingerprint patterns are roughly classified into each category in the second hidden layer