Research Article
A Deep Learning App for Counterfeit Banknote Detection in the WAEMU
@INPROCEEDINGS{10.4108/eai.24-11-2022.2329802, author={Aboudramane Diarra and Tegawend\^{e} F. Bissyande and Pasteur Poda}, title={A Deep Learning App for Counterfeit Banknote Detection in the WAEMU}, proceedings={Proceedings of the 5th edition of the Computer Science Research Days, JRI 2022, 24-26 November 2022, Ouagadougou, Burkina Faso}, publisher={EAI}, proceedings_a={JRI}, year={2023}, month={5}, keywords={counterfeit detection cfa banknotes deep learning convolutional neural network android}, doi={10.4108/eai.24-11-2022.2329802} }- Aboudramane Diarra
Tegawendé F. Bissyande
Pasteur Poda
Year: 2023
A Deep Learning App for Counterfeit Banknote Detection in the WAEMU
JRI
EAI
DOI: 10.4108/eai.24-11-2022.2329802
Abstract
In West Africa, counterfeit CFA banknotes impact the economic growth of states. Because of this scourge, we are witnessing a decline in the purchasing power of the population. However, some hardware kits for detecting counterfeit CFA banknotes are available on the market. These kits are expensive for the actors of the informal sector. For them, these kits are not easily portable and also frequently break down. In this work, we propose an approach based on deep learning for the detection of counterfeit CFA banknotes through an Android application. Furthermore, the proposed approach is a first one in the WAEMU in the field of research for the CFA banknote forgery detection. We use an image dataset of over 4000 genuine and counterfeit ten-thousand CFA banknote for our model training. The images of the banknotes are taken by a smartphone camera. We use the convolutional neural network Alexnet for banknote classification. The accuracy of the training model reaches 99.7% for the detection of counterfeit CFA banknotes.