Classification Of Indonesian Rupiah Currency Using Convolutional Neural Network For Visually Impaired Individuals
Muhammad Auvi Amril, Faqih Hamami, Ahmad Musnansyah · 2024
Individuals with visual impairments face challenges in identifying Indonesian Rupiah currency, affecting their ability to engage in daily transactions. Previous research has proposed various solutions using image recognition technology and artificial neural networks, yet there remains a need for further development, particularly in addressing currency design variations and changes. This study aims to develop a Rupiah currency classification system based on Convolutional Neural Network (CNN) deployed on Android devices to enhance accuracy and responsiveness, while accounting for suboptimal currency conditions. It also explores the impact of hyperparameters on model performance and optimizes the use of Android devices. The model can classify Indonesian rupiah with an accuracy of 0.97 in liv- tests with 70 trials. Therefore, this study is expected to assist visually impaired individuals in conducting transactions.