Fingerprint Feature Extraction with Artificial Neural Network and Image Processing Methods

Abstract

Fingerprints are often claimed as the safest measurable human trait. As one can observe they are commonly used in many different solutions. Nowadays they are applied for instance in financial institutions where clients can confirm their identities with fingerprints. These biometrics are the kind of the password that no one can lost or forget. An approach to extract fingerprints features with artificial neural network and image processing algorithms is presented in this work. Soft computing methods are becoming more popular. It leads to their usage in the human recognition procedures. In the case of the algorithm presented in this paper, the results from neural network are confirmed by Crossing-Number (CN) algorithm. The two-step confirmation ensures more precise results than obtained with each of them separately. The final stage of the algorithm is minutiae extraction and marking them in the analyzed image.