Urdu text in natural scene images

本研究首次推出了一个包含500张自然场景中乌尔都语文字图像的数据集。该数据集由国际伊斯兰大学电气与计算机工程系创建，旨在填补乌尔都语文字检测领域的数据空白。数据集中的图像通过智能手机摄像头采集，涵盖了不同的光照条件，以模拟真实环境。该数据集主要用于训练和测试乌尔都语文字检测模型，特别是在复杂背景下的文字提取。此外，该数据集还为研究乌尔都语文字的识别和理解提供了基础，有助于开发针对乌尔都语地区的导航和自动驾驶技术。

This study presents for the first time a dataset containing 500 Urdu text images captured in natural scenes. Developed by the Department of Electrical and Computer Engineering at the International Islamic University, this dataset aims to fill the existing data gap in the domain of Urdu text detection. The images were collected using smartphone cameras, encompassing a wide range of lighting conditions to simulate real-world scenarios. The primary applications of this dataset are the training and testing of Urdu text detection models, especially for text extraction from complex backgrounds. Additionally, it serves as a foundational resource for research on Urdu text recognition and understanding, and supports the development of navigation and autonomous driving technologies tailored for Urdu-speaking regions.