Software system to detect counterfeit Egyptian currency

Currency counterfeiting is one of the most important sources of economic anxiety among governments because of its significant threat to national economy. The present work aims at implementing a system to efficiently and accurately distinguish counterfeit Egyptian banknotes from genuine ones. The approach consists of a number of components including image processing, image histogram, color moments, characteristics extraction and texture analysis. Case studies use many techniques to construct benchmarks based on dozens of genuine currencies. And then we used measures of proximity to decide whether the tested paper is genuine. These techniques use the approach of exploratory data analysis which utilizes: histogram, mean and median defined ranges, color moments, Euclidean distance, and texture analysis. Texture analysis uses image segmentation to extract three security features, namely, anti-scan, optically variable ink, and latent image. The proposed approach is proved remarkably success in distinguishing between genuine and even the most elaborate counterfeit currency.