Effects of preprocessing and layered neural networks on individual identification

We constructed individual identification system of fingerprint by three-layered neural networks, and investigated the effects of preprocessing and neural networks on the recognition rates of individual identification. The fingerprint images were classified into four directions, then smoothing was done and the neural networks were optimized. From the results, we found that the recognition rates in our system were about 99%, and that the preprocessing and the layered neural networks were useful for individual identification system with the high recognition rates.