STPLS3D

STPLS3D是一个大规模的合成与真实航空摄影测量3D点云数据集，由南加州大学创意技术研究所创建。该数据集包含高质量、丰富注释的点云，覆盖超过17平方公里的城市景观，包括真实和合成环境。数据集创建过程中，利用了开放的地理空间数据源和现成的商业软件包，通过模拟真实环境的无人机飞行模式和重建过程，确保了数据的质量、噪声模式和多样性与真实数据相似。STPLS3D数据集主要用于解决大规模3D点云数据的收集、清洗和标注问题，以及提供精确的语义和实例注释，避免昂贵和耗时的手动标注。

STPLS3D is a large-scale synthetic and real aerial photogrammetric 3D point cloud dataset developed by the Institute for Creative Technologies of the University of Southern California. This dataset provides high-quality, richly annotated point clouds spanning urban landscapes across an area of over 17 square kilometers, encompassing both real-world and synthetic environments. During its development, open geospatial data sources and off-the-shelf commercial software packages are utilized, and real-world unmanned aerial vehicle (UAV) flight patterns and reconstruction workflows are simulated to ensure that the data's quality, noise characteristics and diversity align with those of real-world collected datasets. The STPLS3D dataset is primarily designed to address challenges in large-scale 3D point cloud data collection, cleaning and annotation, while providing precise semantic and instance annotations to eliminate costly and time-consuming manual annotation work.