Currency Identification and Fraudulence Detection System

Abstract

Counterfeit currency is often rumored around the world. Growing up in tandem with technology, there is a huge cycle of making counterfeit notes. Just as it has become easier to make counterfeit currency day by day, it has become harder to detect counterfeit currency. The only way to stop fraud is to use a wide range of counterfeit detection tools or software that are easy to use. There are still many people who cannot afford a fake detection tool or software. So, these tools or software should be free, reliable and effective in terms of accuracy. This report describes a powerful counterfeit currency detection software system for Bangladeshi banknotes. And that which ordinary people can use for free. The main function of this software is to identify the currency and find out whether the currency is genuine or counterfeit. To make this whole process work, deep learning algorithm like Convolutional Neural Network (CNN) has been used to identify the currency and FLANN-based Matcher with the Scale-Invariant Feature Transform (SIFT) algorithm has been used to identify whether the currency is genuine or counterfeit. An input is observed with the help of many training images. It then recognizes the currency with high confidence by analyzing the results of their match. And through this process counterfeit currency can be detected in a very short time. We've put the system into a mobile and web application so everyone can use it quickly and easily.