Fingerprint Recognition Using Discrete Wavelet Transform And Neural Network For Estimation Rotation Region


Fingerprint-based recognition is one of the most important biometric technologies which have drawn a substantial amount of attention recently. This paper proposed a fingerprint recognition adopting a multilayer back propagation neural network as decision stage and using high pass filters (horizontal and vertical sub band the out from 2-D DWT) as estimation rotation region is first stage. The output of this stage is used as input to 1-D DWT (multi-order of low pass decomposition) as feature extraction for fingerprint. Next the normalization for each feature vector is performed to reduce the size of storage and increased the learning speed for the neural network. The evaluation tests were carried out on the proposed algorithm using a database of fingerprint images. A perfect results of recognition (99%) results was achieved by using correlation measurement. we identify the fingerprint by comparing the other fingerprint features with shift position and changed angle to the measurement of performances.