Urdu Handwritten Digits and Characters Dataset

本研究介绍了首个用于乌尔都语手写数字和字符自动识别的数据集，由伊斯兰堡COMSATS大学信息技术与电气工程学院创建。该数据集包含来自900多名不同年龄段个体的样本，涵盖10个数字和40个字符，总计45000张28x28像素的图像。数据集通过扫描手写样本并进行预处理，包括灰度转换、噪声去除和分割，以确保数据质量。此数据集旨在为乌尔都语手写文本的自动识别研究提供基础，解决该领域数据资源匮乏的问题，并推动相关技术的发展。

This study introduces the first dataset for automatic recognition of Urdu handwritten digits and characters, which was created by the Department of Information Technology and Electrical Engineering at COMSATS University Islamabad. This dataset contains samples from over 900 individuals across various age groups, covering 10 digits and 40 characters, with a total of 45,000 28×28 pixel images. The dataset was developed by scanning handwritten samples and performing preprocessing steps including grayscale conversion, noise removal and segmentation to ensure data quality. This dataset aims to provide a foundation for research on automatic recognition of Urdu handwritten text, address the shortage of data resources in this field, and promote the development of related technologies.