Indian Coin Currency Dataset

Published: 9 May 2022| Version 1 | DOI: 10.17632/k49x47v732.1


Currency recognition and classification is one essential task to do. Both paper and coin currency play important role in transactions in everyday life. But provided there are many datasets available for paper currency, and very less datasets are available for coin currency. Coin currency recognition becomes important because even though the amount for which people do coin transactions is small but inaccuracy in recognition can lead to huge loss. Following are the objectives to create this dataset: · Create a dataset of old and new Indian coin currency. · Dataset of good quality images with properly fixed dimensions of 256*256 pixels. · Dataset is in two formats, Yolo and PASCAL(VOC). This image dataset consists of 1907 images, having 4 classes: Re 1 coin, Rs 2 coins, Rs 5 coins, and Rs 10 coins. The images were taken by the mobile phone’s rear camera. They were taken under different conditions like light differences, and different backgrounds, also each class has a different type of coin, for example, a 1re coin has 3 types of different coins based on its design and structure.



Vishwakarma Institute of Information Technology


Artificial Intelligence, Image Processing, Machine Learning