Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning

Banknotes are used worldwide, enabling people to exchange products and services. However, visually impaired and blind (VIB) individuals may face difficulties handling banknotes, especially faded or worn banknotes. Therefore, this research proposed a real-time banknote recognition system using Single...

Full description

Bibliographic Details
Published in:2023 4th International Conference on Artificial Intelligence and Data Sciences: Discovering Technological Advancement in Artificial Intelligence and Data Science, AiDAS 2023 - Proceedings
Main Author: Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
Format: Conference paper
Language:English
Published: Institute of Electrical and Electronics Engineers Inc. 2023
Online Access:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85176544650&doi=10.1109%2fAiDAS60501.2023.10284630&partnerID=40&md5=91a8620dbc5e77f03c65bbb10a0b0ab4
id 2-s2.0-85176544650
spelling 2-s2.0-85176544650
Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
2023
2023 4th International Conference on Artificial Intelligence and Data Sciences: Discovering Technological Advancement in Artificial Intelligence and Data Science, AiDAS 2023 - Proceedings


10.1109/AiDAS60501.2023.10284630
https://www.scopus.com/inward/record.uri?eid=2-s2.0-85176544650&doi=10.1109%2fAiDAS60501.2023.10284630&partnerID=40&md5=91a8620dbc5e77f03c65bbb10a0b0ab4
Banknotes are used worldwide, enabling people to exchange products and services. However, visually impaired and blind (VIB) individuals may face difficulties handling banknotes, especially faded or worn banknotes. Therefore, this research proposed a real-time banknote recognition system using Single Shot Detector (SSD) Mobilenetv2 algorithm. The SSD Mobilenetv2 algorithm is efficient, fast, and suitable for low-end devices, making it a suitable technique for this research. A dataset of six Malaysian banknote images (RM1,5,10,20,50, and 100) was augmented and pre-processed was used in this research. Several experiments were conducted to optimize the model's performance using Python, TensorFlow, and Keras. The optimized model achieved 77% accuracy with an average speed suitable for mobile device implementation, using an input size of 320\times 320, 80:20 data split, 32 batch size, and 30,000 training steps. The application was able to detect the notes under various real-time settings like illumination, occlusion, or even half-folded notes, with an audible output voice. Subsequently, further development is necessary to enhance the accessibility and user-friendliness of the application. Increasing the dataset's size and variation of banknotes would help improve the accuracy, and adding vibration alerts when a note is detected would also be beneficial. These improvements empower visually impaired individuals, enhancing their independence and interaction with the world. © 2023 IEEE.
Institute of Electrical and Electronics Engineers Inc.

English
Conference paper

author Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
spellingShingle Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
author_facet Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
author_sort Hanafiah N.I.M.; Mahmud Y.; Bin Hussin M.H.; Kamarudin S.N.K.
title Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
title_short Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
title_full Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
title_fullStr Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
title_full_unstemmed Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
title_sort Malaysian Banknote Recognition App for the Visually Impaired Using Deep Learning
publishDate 2023
container_title 2023 4th International Conference on Artificial Intelligence and Data Sciences: Discovering Technological Advancement in Artificial Intelligence and Data Science, AiDAS 2023 - Proceedings
container_volume
container_issue
doi_str_mv 10.1109/AiDAS60501.2023.10284630
url https://www.scopus.com/inward/record.uri?eid=2-s2.0-85176544650&doi=10.1109%2fAiDAS60501.2023.10284630&partnerID=40&md5=91a8620dbc5e77f03c65bbb10a0b0ab4
description Banknotes are used worldwide, enabling people to exchange products and services. However, visually impaired and blind (VIB) individuals may face difficulties handling banknotes, especially faded or worn banknotes. Therefore, this research proposed a real-time banknote recognition system using Single Shot Detector (SSD) Mobilenetv2 algorithm. The SSD Mobilenetv2 algorithm is efficient, fast, and suitable for low-end devices, making it a suitable technique for this research. A dataset of six Malaysian banknote images (RM1,5,10,20,50, and 100) was augmented and pre-processed was used in this research. Several experiments were conducted to optimize the model's performance using Python, TensorFlow, and Keras. The optimized model achieved 77% accuracy with an average speed suitable for mobile device implementation, using an input size of 320\times 320, 80:20 data split, 32 batch size, and 30,000 training steps. The application was able to detect the notes under various real-time settings like illumination, occlusion, or even half-folded notes, with an audible output voice. Subsequently, further development is necessary to enhance the accessibility and user-friendliness of the application. Increasing the dataset's size and variation of banknotes would help improve the accuracy, and adding vibration alerts when a note is detected would also be beneficial. These improvements empower visually impaired individuals, enhancing their independence and interaction with the world. © 2023 IEEE.
publisher Institute of Electrical and Electronics Engineers Inc.
issn
language English
format Conference paper
accesstype
record_format scopus
collection Scopus
_version_ 1809678021100044288