Current - Issue

One of the most important assets of our country is its bank currency. However, to create discrepancies in the financial market, criminals introduce counterfeit notes that closely resemble the original ones. During the demonetization period, a significant amount of fake currency was observed circulating in the market. Generally, it is quite challenging for a person to distinguish between a forged note and a genuine one, despite various identification parameters, as many features of counterfeit notes are similar to those of authentic currency. Differentiating between fake banknotes and real ones is a difficult task. Therefore, there is a need for an automated system that can be implemented in banks or ATM machines. To create such a system, an efficient algorithm must be designed to determine whether a banknote is genuine or counterfeit, as fake notes are crafted with high precision. In this paper, we utilize the Convolutional Neural Network (CNN) algorithm on a dataset available from the UCI Machine Learning Repository for the purpose of bank currency authentication. We have applied machine learning algorithms and evaluated their performance based on various quantitative analysis parameters.