Authors: Redna A. Almutlaq; Daliyah S. Aljutaili; Suha A. Alharbi; Dina M. Ibrahim
Addresses: Department of Information Technology, College of Computer, Qassim University, Buraydah, 51452, Saudi Arabia ' Department of Information Technology, College of Computer, Qassim University, Buraydah, 51452, Saudi Arabia ' Department of Information Technology, College of Computer, Qassim University, Buraydah, 51452, Saudi Arabia ' Faculty of Engineering, Computers and Control Engineering Department, Tanta University, Tanta, 31733, Egypt
Abstract: Most commercial transactions are still done using physical currencies. Detection of fake currency can improve the reliability of ATMs and counting machines in order to ensure proper maintenance operations and confirm the value and authenticity of currency. Many paper currencies or banknotes are exposed to the presence of some problems, such as folding, rolling, and wrinkling. In this research, we study and analyse the impact of applying one of the currency recognition algorithms, which is speeded up robust scale invariant feature transform (SR-SIFT) algorithm in detecting the currency paper in the normal presentation and also in any troubles like rolled, wrinkled, or folded. The results show that the SR-SIFT algorithm can usefully recognise the currencies in a different situation with high accuracy.
Keywords: currency recognition; speeded up robust scale invariant; SR-SIFT; folded banknotes; wrinkled banknotes; rolled banknotes; normal banknotes; commercial transactions; currency detection; high accuracy.
International Journal of Data Science, 2020 Vol.5 No.2, pp.151 - 159
Received: 25 Apr 2020
Accepted: 06 Jul 2020
Published online: 11 Dec 2020 *