Yemeni Currency Recognition Dataset

Published: 17 December 2025| Version 2 | DOI: 10.17632/s56nbwsytx.2
Contributor:
MAJED MOLHI

Description

This dataset contains real-world images of Yemeni banknotes collected for research in currency recognition and computer vision. The images were captured using smartphone cameras under diverse conditions, including variations in illumination, background, orientation, scale, and partial occlusions. The dataset includes six Yemeni Riyal denominations: 50, 100, 200 (old), 250, 500, and 1000 YER. Each denomination is organized in a separate folder. The dataset consists of a total of 1,691 RGB images. This dataset is intended to support tasks such as image classification, object detection, and evaluation of currency recognition systems, including assistive applications for visually impaired users. It is designed to facilitate reproducible research and testing of deep learning models under realistic acquisition conditions.

Files

Steps to reproduce

1. Download the dataset files. 2. Organize images by denomination folders. 3. Use the dataset directly for training or evaluation of currency recognition or detection models. 4. No additional preprocessing is required.

Institutions

Ibb University

Categories

Computer Vision, Object Detection, Machine Learning, Image Classification

Licence