Simultaneous detection of multiple counterfeit banknotes using Raman spectroscopy: A case study of ₦500 and ₦1000 notes

Abstract

A bundle of ₦500 and ₦1000 notes were analyzed using Raman spectroscopy to simultaneously detect and localize multiple counterfeit  notes mixed with genuine ones in the bundle. The research applied machine learning techniques, specifically Multiview non-negative  matrix factorization (Mv-NMF) and 2D correlation analysis, to detect and assess the spectral similarity between the bundle contents and a  reference genuine note, identifying individual spectra as either genuine or counterfeit. Furthermore, an approach was provided for the spatial localization of each sample bank note within a bundle, offering a detailed visualization of all banknotes in the bundle  simultaneously. This research showcases the potential of Raman spectroscopy in molecular forensic analysis of a bundle for detecting  and identifying the spectral signatures of multiple bank notes with a single banknote's security features, offering a viable alternative to  existing individual banknote analysis and detection methods. The research may also be applicable to any security documents with a  unique substrate paper that can be bundled together. Our findings offer significant benefits to financial authorities, legal institutions, and  intelligence organizations around the world for the smart identification of multiple counterfeited notes.