Identity recognation via fingerprint analysis by neural networks

Today, it is an undeniable fact that fingerprints play quite important role in the areas of security and forensic investigation for authentication. In this work, a set of fingerprints are recorded, classified and analyzed using artificial neural network (ANN) structure. For this purpose, fingerprints taken from both internet database and real individuals are used together, and the attributes of these fingerprint images are extracted after processing. Beforementioned attributes are used in the statistical analysis and the resulted data are trained and/or tested by artificial neural networks.As a result of not being perfect, the increase in the number of fingerprint records used in the train set of the ANN reduces the success rate of the application, However, the success ratio at the end of this study is obtained between 90-93%. Compared with the similar studies published so far, it can be said that this study has the acceptable success level