Recognition of Brazilian Banknotes Using Artificial Intelligence to Assist Visually Impaired Individuals

Authors

  • Karen de Almeida UNESP
  • Almir Olivette Artero UNESP
  • Danilo Medeiros Eler UNESP
  • Maurício Araújo Dias UNESP
  • Danillo Roberto Pereira UNESP
  • Francisco Assis da Silva UNOESTE

DOI:

https://doi.org/10.5335/rbca.v17i1.15934

Keywords:

Accessibility, Brazilian Banknote Recognition, Convolutional Neural Networks

Abstract

Recognizing Brazilian Real banknotes remains a challenge for visually impaired individuals. This work presents a proposal that employs a convolutional neural network to assist visually impaired people in recognizing Real banknotes using a smartphone with a camera. The implementation provides an application called Banknotes Recognition, which performs the recognition process swiftly by capturing the banknote image with the smartphone's camera. Tests conducted with a set of 66 images of Real banknotes showed an accuracy of 100%, with all banknotes correctly classified even under unfavorable conditions, such as folded notes, complex backgrounds, various lighting conditions, and worn-out notes.

Downloads

Download data is not yet available.

Published

2025-05-23

Issue

Section

Original Paper

How to Cite

[1]
2025. Recognition of Brazilian Banknotes Using Artificial Intelligence to Assist Visually Impaired Individuals. Brazilian Journal of Applied Computing. 17, 1 (May 2025), 23–32. DOI:https://doi.org/10.5335/rbca.v17i1.15934.