Fakes are a major threat to India’s economy hence the need for a solution that will help detect these materials easily and accurately. The detection of counterfeit currency is a major concern for almost all countries, and this research focuses on developing a MATLAB-based counterfeit currency detection system which evaluates a true image of the Indian Currencies with the help of image processing and pattern recognition technique. Benefiting from MATLAB’s powerful image processing resources, the system conducts the necessary preprocessing, feature extraction and classification of vital security elements of currency, such as watermarks, security threads and micro-lettering which play an important role in identifying the genuine currency from the counterfeit. The specific characteristics of edge and texture are statistically and geometrically calculated, and the normal and high-resolution light conditions are at high accuracy with varying resolutions. In order to determine a distinction between actual and fake notes, support vector machine (SVM) classifiers are used. By validating this MATLAB solution, it has been determined to be effective as an easy to use, robust and customizable software that has the potential to work in numerous operations within banking and retail and prevent the spreading of counterfeit money.
Read full abstract