Empowering the visually impaired: A YOLO-v5 CNN-based android app for currency recognition and object detection

Abstract

This paper presents an Android mobile Application (app) that assists people with visual impairments in currency recognition and general object detection. Our project's goal is to assist visually impaired individuals by offering them a practical alternative with a cost-effective solution. The YOLO-v5 CNN model-based currency note detection and object recognition app is proposed for this purpose, and we find it is fast and accurate. The app can identify rupee notes with denominations of 10 new, 10 old, 20 new, 20 old, 50 new, 50 old, 100 new, 100 old, 200, 500, 2000, and general objects such as mobile phones, laptops, chair, water bottle, vehicles, persons, book, door, watch etc.