Building Footprints 2022

Building polygons across the 1,199 sq mi project area was captured through a semi-automated workflow using high-resolution lidar and ortho aerial imagery data collected in 2022. Automated algorithms first classified ground in the lidar point cloud and were reviewed for accuracy. After finalization of the ground classification, additional algorithms classified above ground features, such as buildings, vegetation, rails, etc. Building classified points were clustered and converted to polygon vectors across the project area. The polygons produced from the described automated process(es) are visually assessed for gross errors and/or significant misclassifications referencing the existing 2022 orthoimagery and County-provided source material (e.g., countywide parcel information). The only corrective action taken was the splitting of single large polygons into multiple polygon features by interpreting the intersection of the parcel information, or the capture of a missing feature (structure). Each building polygon was lastly assigned an elevation value based on the highest lidar point in the interior of its respective polygon. Elevation values were reviewed for accuracy. All building polygons have attributes including square footage, collection year of lidar and imagery, ground elevation, highest lidar point, and height above ground.本数据集涵盖的1199平方英里项目区域内的建筑多边形，依托2022年采集的高分辨率激光雷达（lidar）与正射航空影像（ortho aerial imagery）数据，通过半自动化工作流完成提取。自动化算法首先对激光雷达点云进行地面分类，并对分类结果开展精度复核。完成地面分类作业后，后续算法对地面以上地物（如建筑、植被、轨道等）进行分类。已被分类为建筑的点云将被聚类，并转换为项目区域内的多边形矢量要素。通过上述自动化流程生成的多边形，将参照2022年正射航空影像与县级提供的源数据（如全县宗地信息）开展目视评估，以排查严重错误或显著误分类问题。本次仅采取两类修正操作：一是通过解析宗地信息的交集，将单个大型多边形拆分为多个多边形要素；二是补全缺失的地物（建筑结构）。最后，为每个建筑多边形分配基于其对应多边形内部最高激光雷达点的高程值，并对高程值的精度进行复核。所有建筑多边形均包含以下属性：建筑面积、激光雷达与影像的采集年份、地面高程、最高激光雷达点高程，以及相对地面高度。