Dataset of Indian (Rupees) and Thai (Baht) Banknotes with Annotations

Published: 16 July 2021| Version 1 | DOI: 10.17632/2kfz5yc7pt.1


Currency recognition and classification is a very challenging task for visually impaired people. In the daily routine, correct and fast currency recognition can help them to do the financial transactions without any help. With this objective we have created the Indian (Rupees) and Thai (Baht) banknotes dataset with annotations. This dataset consists of total 3000 (2000 images of Indian banknotes and 1000 images of Thai banknotes) high quality images. Mobile phone's rear camera was used to take the images of Indian and Thai banknotes. The Indian banknote dataset consists of 10 classes namely 10 New, 10 Old, 20, 50 New, 50 Old, 100 New, 100 Old, 200, 500, 2000 and Thai banknote consists of 5 classes namely 20, 50, 100, 500, and 2000. The images were taken at the different backgrounds and in different lighting conditions, in the cluttered background, and also the images of folded banknotes are taken. The images are annotated in the YOLO (You Only Look Ones) format. The dataset consists of "Training" and "Validation" images in 80:20 ratio.



Computer Vision, Object Detection, Object Recognition, Machine Learning, Convolutional Neural Network, Deep Learning