CartoMark

CartoMark是一个大规模的基准数据集，由中国矿业大学环境与空间信息学院等机构合作创建，旨在通过机器智能技术提升地图模式识别和内容检索的能力。该数据集包含约20000张来自互联网、数据存储库和社交媒体的地图，覆盖多种格式和内容，如地图文本标注识别、场景分类、超分辨率重建和风格转换。创建过程涉及数据收集、清洗和标注，确保数据质量和多样性。数据集的应用领域广泛，包括文化叙事、导航服务和基于位置的服务，旨在解决地图数据的有效利用和深度学习在复杂地图内容识别中的应用问题。

CartoMark is a large-scale benchmark dataset co-developed by institutions including the School of Environment and Spatial Informatics, China University of Mining and Technology, aiming to enhance map pattern recognition and content retrieval capabilities via machine intelligence technologies. This dataset contains approximately 20,000 maps sourced from the Internet, data repositories and social media, covering diverse formats and content types such as map text annotation recognition, scene classification, super-resolution reconstruction and style transfer. Its creation process involves data collection, cleaning and annotation to ensure data quality and diversity. The dataset has a wide range of application fields including cultural narration, navigation services and location-based services, and is designed to address the issues of effective utilization of map data and the application of deep learning in complex map content recognition.