Counterfeit Malaysian Banknotes Detection Using Discrete Wavelet Transform

Abstract

The number of counterfeit banknotes has increased year by year, due to the enhanced developments in printing technology. Due to technological advancement, counterfeit banknotes can pass the physical feature and chemical property-based counterfeit banknotes detection system undetected. The end-user least accepts the fake detection tools because of poor accuracy, unavailability and expensive. As a result, counterfeit banknotes have become a major issue for most countries in the world, including Malaysia. This paper proposes a method for the Malaysian counterfeit banknote detection machine to use the Haar wavelet transform as the featured extraction technique. The features were then classified by using the K-Nearest Neighbour (KNN) and Support Vector Machine (SVM) classifiers. The proposed models would produce 85% and 50% accuracies respectively for KNN and SVM. In the future, this system can be enhanced into real-time banknotes detection