Data and <strong>coding</strong> used in paper entitled <strong>"MIU: Deep Embedded Building Cluster Model of Urban Functional Zoning"</strong>

Data and <strong>coding</strong> used in paper entitled <strong>"MIU: Deep Embedded Building Cluster Model of Urban Functional Zoning".The compressed package contains 6 folders.</strong> <strong>Building Footprint: Building vector data were used to extract geometric and compactness featrues.</strong> <strong>Google Earth Image: VHR images were applied to extract spectral and textural features.</strong> <strong>Luojia 1-01 Nighttime Light Image: Nighttime data were used to extract brightness features.</strong> <strong>OSM Street:OSM road networks were used to extract location features.</strong> <strong>POI of Study Area:POI data were used to generate labels for training the Word2Vec model.</strong> <strong>Python Code:DEC code was used to process the cluster for generating the MIU; Word2Vec code was used train the Word2Vec model.</strong>**本数据集为论文《MIU：面向城市功能分区的深度嵌入式建筑聚类模型》所使用的数据与编码代码。该压缩包包含6个文件夹。** **建筑足迹（Building Footprint）**：采用建筑矢量数据提取几何特征与紧凑度特征。 **谷歌地球影像（Google Earth Image）**：采用超高分辨率（Very High Resolution, VHR）影像提取光谱特征与纹理特征。 **珞珈一号夜间灯光影像（Luojia 1-01 Nighttime Light Image）**：采用夜间灯光数据提取亮度特征。 **OSM道路网（OSM Street）**：采用开放街道地图道路网络提取位置特征。 **研究区兴趣点（Point of Interest, POI）数据**：利用POI数据生成用于训练Word2Vec模型的标签。 **Python代码（Python Code）**：包含用于聚类处理以生成MIU的DEC代码，以及用于训练Word2Vec模型的代码。