Utilizing Deep Learning with CNN Model for Precise Identification of Bangladeshi Currency Notes to Aid Visually Impaired People

Visually impaired people need technological support to accurately predict and analyze authentic currency notes. There is a requirement to create a software application capable of automatically detecting and recognizing different types of currency notes. In today's advancing world, automated currency recognition systems play an integral role, particularly when it comes to identifying paper currency. Machines often struggle to identify and recognize currencies in circulation, especially when the currency notes are damaged. Nevertheless, employing Deep Learning (DL) techniques with Convolutional Neural Networks (CNN) can serve as an effective solution for visually impaired individuals to accurately identify currency notes. Considering these thoughts, this paper presents a forecasting technique for Bangladeshi currency printed on paper and introduces an enhanced procedure for accurately identifying currencies. We then propose a deep neural network method employing CNN, which improves the effectiveness of currency authentication by offering increased precision, high speed, and efficiency. With the aid of the RestNet50 model, 99 percent accuracy was achieved. This method performs autonomously, requiring minimal human intervention. Additionally, we utilize realtime data to evaluate the effectiveness and potential of the presented approach by comparing it to existing methods.