Building Segmentation and Modelling from Space-Borne and Aerial Imagery

Abstract. Accurate 3D building reconstruction is essential for urban planning, disaster management, and environmental applications. However, current methods often struggle to achieve geometric precision and topological consistency, particularly when when processing satellite or aerial imagery. This paper presents a comprehensive workflow that addresses these challenges, enabling the generation of multiple outputs—including digital surface models (DSMs), digital terrain models (DTMs), true-orthophotos, 2D building segments, and vectorized 3D LoD-2 building models. Our approach leverages very high-resolution (VHR) imagery to derive precise DSM and DTM data, which are used in conjunction with orthorectified imagery to accurately segment buildings and delineate roof planes. By focusing on planar building components and employing robust vectorization techniques, our workflow ensures consistent 3D model construction while avoiding the challenges of fine-detail extraction.Validated on diverse urban datasets, our method demonstrates high accuracy, scalability, and potential to advance building reconstruction workflows in remote sensing, contributing significantly to geospatial and environmental research.