CMAB
收藏资源简介:
CMAB数据集由清华大学创建,是中国首个全国范围的多属性建筑数据集,涵盖了3667个自然城市,总面积达213亿平方米。该数据集通过集成多源数据,如高分辨率Google Earth影像和街景图像,生成了建筑的屋顶、高度、功能、年龄和质量等属性。数据集的创建过程结合了地理人工智能框架和机器学习模型,确保了数据的高准确性。CMAB数据集主要应用于城市规划和可持续发展研究,旨在提供详细的城市3D物理和社会结构信息,支持城市化进程和政府决策。
The CMAB dataset, developed by Tsinghua University, is China's first national-scale multi-attribute building dataset, covering 3,667 natural cities with a total area of 21.3 billion square meters. This dataset integrates multi-source data including high-resolution Google Earth imagery and street view images to generate building attributes such as roof characteristics, height, function, construction age, and quality. Its development process combines geographic artificial intelligence (GeoAI) frameworks and machine learning models, ensuring high data accuracy. Primarily applied in urban planning and sustainable development research, the CMAB dataset aims to provide detailed 3D physical and social structural information of cities to support urbanization processes and government decision-making.




