EUVIMEDEUVIMED entdecken
Crossref

Deep Neural Network Based Fake Currency Detection Using FPGA for Real Time Verification

K. Usha, M.Tulasi Venkat, K.Shiva Karthik, M. Praveen Kumar

International Journal of Recent Development in Engineering and Technology · 22.04.2026 · 15 (4) · 465-474

DOI: 10.54380/ijrdet0426_75 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Recent Development in Engineering and Technology, online erschienen am 22.04.2026. Verfasst von K. Usha und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Fraudulent notes detection is essential for safeguarding the counties financial system, and will help in preventing the fake currency circulation by upholding the reliable error-free transaction process. This paper presents a stacking ensemble method by which we can have multiple models and train each of them with one particular feature that we want to extract from the currency and by employing a metamodel to learn the best way to combine will be a better approach to solve this issue. The recommended strategy adopts a squeezenet lite model and a simplified 6-layer CNN structure for lightweight structure and compatibility for hardware oriented implementation. Squeezenet-lite model consist of a initial input convolution layer followed by the 9 fire module each consists of the squeeze and expand layers which has 1✕1 convolutional filters, enabling efficient feature extraction. The features considered for the training are RGB images & Image Quality Features(IQF) for acquiring both color-based and quality-related characteristics of currency notes. The dataset contains 200 and 500 denominations used for training and evaluation. Squeeznet model will analyze the RGB images and the IQF features are processed by 6-layer CNN. Based on these two model outputs the Support Vector Machine learns the thresholds of the classification of the currency will provide better performance.The models are trained in the MATLAB and targeted for FPGA implementation to enable faster inference and real-time counterfeit detection.

Zitieren

APA

Usha, K., Venkat, M., Karthik, K., & Kumar, M. P. (2026). Deep Neural Network Based Fake Currency Detection Using FPGA for Real Time Verification. International Journal of Recent Development in Engineering and Technology, 15(4), 465-474. https://doi.org/10.54380/ijrdet0426_75

Vancouver

Usha K, Venkat M, Karthik K, Kumar MP. Deep Neural Network Based Fake Currency Detection Using FPGA for Real Time Verification. International Journal of Recent Development in Engineering and Technology. 2026;15(4):465-474. doi:10.54380/ijrdet0426_75

BibTeX
@article{Usha2026,
  author = {Usha, K. and Venkat, M.Tulasi and Karthik, K.Shiva and Kumar, M. Praveen},
  title = {Deep Neural Network Based Fake Currency Detection Using FPGA for Real Time Verification},
  journal = {International Journal of Recent Development in Engineering and Technology},
  year = {2026},
  volume = {15},
  number = {4},
  pages = {465--474},
  doi = {10.54380/ijrdet0426_75}
}
Kostenlos starten →