Deep Learning based Money Detection System for Visually Impaired Person
Aniket Jawale, Kanchan Patil, Aniruddha Patil, Siddhesh Sagar, Akash Momale · 2023
In recent times, visually challenged people are facing challenges in completing their daily tasks. One of the issues is to use bank money to make purchases and also to deal with financial fraud. In light of this, this study has developed a novel system to automatically identify the currency and provide output in audio format. The proposed system is composed of various modules. Using a camera, the data is collected and given as input to the first module. Following pre-processing, the deep learning model will determine whether the currency is genuine or not, and if it is, it will classify the currency. In order to do this, this study has examined various deep learning techniques. For this, many deep learning object identification and classification algorithms are examined before choosing the one that performs the task the best. In the following module, NLP is utilized to extract text from the currency and create the text. The created text will be converted into audio with the aid of the Text-to-Speech module, and the audio output will be provided to blind users via a Raspberry Pi connected audio device.