Robust Designs for Fingerprint Feature Extraction CNN with Von Neumann Neighborhood

Abstract

The cellular neural/nonlinear network (CNN) is a powerful tool for image and video signal processing, robotic and biological visions. The robust designs for CNN templates are important issue for the practical applications of the CNN. The fingerprint feature extraction (FFE) CNNs are two kinds of CNNs, which are able to extract the endings and bifurcations in patterns, two important features in a fingerprint image. This paper establishes two theorems for designing the robustness templates of these two kinds of FFE CNNs respectively. These two theorems provide the template parameter inequalities to determine parameter intervals for implementing the corresponding functions. Simulation result shows the effectiveness of the proposed methodology.