LBPNET: Detection of Fake Face Currency through LBP-Based Convolutional Neural Networks

The escalating sophistication of counterfeit face currency demands innovative solutions for robust detection. In this study, we propose LBPNET, a novel approach employing Local Binary Pattern (LBP) feature extraction and Convolutional Neural Networks (CNNs) to address the challenges posed by realistic fake currencies. LBP descriptors are extracted from currency images, capturing subtle patterns invisible to the human eye. These descriptors serve as inputs for a CNN, named LBPNET, trained to discern genuine from counterfeit face currencies. The model undergoes a comprehensive training phase, incorporating diverse currency images, to establish a generalized and adaptive detection system. In testing, new currency images are evaluated using the trained LBPNET to identify the presence of fake or non-fake currency, showcasing the model's adaptability to emerging counterfeit variations. This approach integrates machine learning and image processing to enhance detection capabilities, offering a promising solution for the evolving landscape of counterfeit face currency across forensic and social media domains.