Machine-learned 3D Building Vectorization from Satellite Imagery

TL;DR

This paper introduces a machine learning pipeline that automatically reconstructs and vectorizes 3D building models from satellite imagery, combining deep learning and vectorization techniques for detailed urban modeling.

Contribution

It presents a novel integrated approach using cGANs and semantic segmentation for accurate 3D building reconstruction from single satellite images.

Findings

Achieves state-of-the-art accuracy on large-scale satellite datasets

Effectively filters non-building objects and refines building shapes

Generates detailed 3D building models with height and roof polygons

Abstract

We propose a machine learning based approach for automatic 3D building reconstruction and vectorization. Taking a single-channel photogrammetric digital surface model (DSM) and panchromatic (PAN) image as input, we first filter out non-building objects and refine the building shapes of input DSM with a conditional generative adversarial network (cGAN). The refined DSM and the input PAN image are then used through a semantic segmentation network to detect edges and corners of building roofs. Later, a set of vectorization algorithms are proposed to build roof polygons. Finally, the height information from the refined DSM is added to the polygons to obtain a fully vectorized level of detail (LoD)-2 building model. We verify the effectiveness of our method on large-scale satellite images, where we obtain state-of-the-art performance.