Banknotes counterfeit detection using deep transfer learning approach

Azra Yildiz, Ali Abd Almisreb, Šejla Dzakmic, Nooritawati Md Tahir, Sherzod Turaev, Mohammed A. Saleh

Research output: Contribution to journalArticlepeer-review


Financial institutions in Bosnia and Herzegovina have adopted and implemented various security measures in preventing production of counterfeited banknotes as well as security measures in detecting counterfeited banknotes. Despite that, there are still many cases of money counterfeiting being reported in this specific country. This happen due to various reasons for instance not thorough attention given to the appearance of the banknote during money transactions. Based on 2017 and 2018 annual report by the Central Bank of Bosnia and Herzegovina, there were 691 and 535 registered cases of banknotes counterfeiting for the convertible mark (KM). The most denomination's banknotes as counterfeited were 100 KM followed by 20 KM and 50 KM. Hence this study proposed deep learning technique in detecting the counterfeited BAM banknotes utilizing CNN models. Initially results showed that the proposed method could be used for counterfeit money detection with 99.88% using VGG16 as the highest accuracy.

Original languageEnglish
Pages (from-to)8115-8122
Number of pages8
JournalInternational Journal of Advanced Trends in Computer Science and Engineering
Issue number5
Publication statusPublished - Sep 1 2020


  • AlexNet
  • Banknotes
  • Counterfeit
  • Deep Transfer Learning
  • GoogLeNet
  • VGG16

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Electrical and Electronic Engineering


Dive into the research topics of 'Banknotes counterfeit detection using deep transfer learning approach'. Together they form a unique fingerprint.

Cite this