Banknote simulator for aging and soiling banknotes using Gaussian models and Perlin noise. (15th December 2019)

• Record Type: Journal Article
• Authors: Baek, Sangwook; Lee, Sanghun; Choi, Euison; Baek, Yoonkil; Lee, Chulhee
• Is Part Of: Expert systems with applications. Volume 137(2019)
• Page Start: 405
• Page End: 419
• Publication Date: 2019-12-15
• Subjects: Banknote aging -- Banknote fitness -- Multi-currency recognition -- Banknote validation -- Gaussian model -- Perlin noise; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2019.07.013
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A banknote aging simulator for generating aged banknotes from new ones is proposed. The proposed simulator allows ATMs to better handle aged and soiled banknotes. With the artificially aged banknote images, classification score improved by 27%. The proposed algorithm can substantially reduce maintenance costs. Abstract: In this paper, we propose a banknote aging simulator for generating aged banknotes from new banknotes. When new banknotes begin to be circulated, automatic teller machines (ATMs) are trained with these new banknotes. However, the ATM operations become more erroneous after several months as the banknotes start aging. Also, the process of updating ATM software for aged banknotes is costly and time-consuming. In this paper, we developed a simulator for banknote aging that uses the Gaussian and Perlin noise models. The proposed simulator allows ATMs to better handle aged and soiled banknotes without updating ATM software. The experimental results show that the classification and validation performance of aged and soiled banknotes improved noticeably when the classification and validation algorithms were designed using the proposed simulator for new banknotes.