Person Identification Biometric Systems from Local Finger-Knuckle Prints Based on Deep Learning

Abstract

Biometrics involves automatically identifying individuals using their physiological or be- havioral characteristics. Multimodal biometric systems integrating multiple modalities recognition methods are considered the optimal solution for accurate identification. This research focuses on using knuckle prints for biometric identification and evaluates the performance of a proposed CNN model. The study explores different feature extraction techniques and classification methods, aiming to determine the most effective combination of knuckle prints for identification.