京沪津穗百米高楼时空数据集（天津）

京沪津穗百米高楼时空数据集（天津）是综合应用测绘地理信息内外业相结合的方式，综合应用“3S”技术（地理信息系统（GIS）、遥感（RS）、全球导航卫星系统GNSS），在全面调查获取天津市房屋现状的基础上，针对超高层建筑类别加工、整理的专项数据集。数据全部为天津市测绘院有限公司自主加工生产。主要包括以下几个关键点过程。①数据准备。以天津市测绘院有限公司自行采集处理得到的天津市最现势的遥感影像底图数据和1:2000地形图数据为基础，以天津市实景三维数据为辅助，利用专业地理信息系统软件，对天津市建筑、特别是高度在100m以上的超高建筑进行精确基底范围提取，以街道、乡镇行政区划为单位，将整个天津市划分为多个普查作业区域，为后续实地调查做准备。②内业数据加工及核查调查底图制作。以准备的最现势的遥感影像数据为底图，通过实景三维模型对建筑高度进行筛查，得到符合基本条件高度的高层建筑三维数据和影像数据，辅以历史多期遥感影像和房屋建筑、道路交通、地铁、绿化等专题数据，对本次百米高度目标数据的建筑年代、到地铁站点及主干道等信息进行加工处理，内业处理初步成果经融合后按区域切片，安装到智能终端调查核查设备中。③实地调查核实。公司组织外业调查人员，通过公司自行研发的天测云图智能终端APP软件，对已有基本信息进行系统梳理整理的同时，同步开展实地信息补充调查采集与核实工作，包括目标建筑的使用类型、名称、所属聚集区、地址、地上层数、主体高度等内容。④数据加工整理。采用天津市测绘院有限公司自研的内网生产作业平台，平台包括48个算力节点，CPU核数达384核，超400GB GPU，达2PB存储资源。结合基础地理信息数据对最终调查核实的数据做标准化处理，对属性信息与空间信息进行关联处理，形成矢量数据集、超高层建筑物数据表格，并按照统一式样分割整理高层建筑影像数据，最后再此基础上融合基础地理信息数据编制分析报告。

The Tianjin Component of the Beijing-Shanghai-Tianjin-Guangzhou 100m+ High-Rise Building Spatio-Temporal Dataset is a specialized dataset processed and organized for super-tall building categories, based on a comprehensive survey of the current status of buildings in Tianjin. It integrates both indoor office and field surveying and mapping geographic information workflows, and comprehensively applies 3S technologies, namely Geographic Information System (GIS), Remote Sensing (RS), and Global Navigation Satellite System (GNSS). All data in this dataset was independently produced and processed by Tianjin Surveying and Mapping Institute Co., Ltd.

The main workflow includes the following key stages:
① Data Preparation: Based on the most up-to-date remote sensing image base map data and 1:2000 topographic map data of Tianjin collected and processed independently by Tianjin Surveying and Mapping Institute Co., Ltd., supplemented by Tianjin's real-scene 3D data, professional geographic information system software was used to accurately extract the base outlines of buildings across Tianjin, especially super-tall structures with a height of 100 meters or above. The entire municipality of Tianjin was divided into multiple census operation zones based on sub-district and township administrative divisions, in preparation for subsequent field surveys.
② Indoor Data Processing and Production of Verification Survey Base Maps: Taking the prepared most up-to-date remote sensing image data as the base map, real-scene 3D models were used to screen building heights to obtain 3D data and image data of high-rise buildings meeting the basic height criteria. Supplementary with historical multi-temporal remote sensing images and thematic data including buildings, road traffic, subway systems, green spaces and others, information such as the construction age of the target 100m+ height buildings, their distances to subway stations and main roads, etc., was processed. After the preliminary indoor processing results were fused, they were sliced by zone and imported into intelligent terminal survey and verification equipment.
③ Field Survey and Verification: The company organized field survey personnel to use the self-developed Tiance Yuntu intelligent terminal APP software to systematically sort out the existing basic information, while simultaneously conducting on-site supplementary surveys, data collection and verification work. Collected information includes the usage type, official name, affiliated cluster area, physical address, above-ground floor count, main structural height, and other relevant details of the target buildings.
④ Data Processing and Organization: The self-developed intranet production operation platform of Tianjin Surveying and Mapping Institute Co., Ltd. was adopted for this stage. The platform is equipped with 48 computing power nodes, totaling 384 CPU cores, over 400GB of GPU resources, and 2PB of storage capacity. Combined with basic geographic information data, the finally surveyed and verified data was standardized, and attribute information was associated with spatial information to generate vector datasets and super-tall building data tables. High-rise building image data was segmented and organized in accordance with a unified format. Finally, based on the above work, the basic geographic information data was fused to compile an analysis report.