Currency Detection Application for Visually Impaired using CNN
Richa Gupta, Amit Kr. Gupta, Aman Chauhan, Gaurav Bisht, Vikas Tripathi · 2023
A visually challenged person may find it challenging to distinguish between various denominations of currency. For those who are blind, this process is still challenging despite the fact that the majority of Indian currency is written in unique characters. Portable solutions for isolation have emerged as a result of a shortage of available tools. By holding paper money in front of a camera, the methodology used mainly aims to develop an android based application which will enable those who are blind to see Indian rupees. This project is based on actual applications from the real world and uses machine learning and Android development techniques. The text-to-speech technology in the Android app reads the user's value description before turning text into speech. To take advantage of the capabilities of machine learning, we employ technologies like ResNet and MobileNetV2 as well as libraries like Keras, TensorFlow, Fastai, and PyTorch. The application's backend uses a variety of technologies, including machine learning models, Python, and different libraries. The java concept and development technique are used for the latter. Overall, they are incorporated into a very user-friendly platform, making it simple to use and use in our daily lives.