Reconstructing Building Height from Spaceborne TomoSAR Point Clouds Using a Dual-Topology Network
Zhaiyu Chen, Yuanyuan Wang, Yilei Shi, Xiao Xiang Zhu
TL;DR
The paper tackles accurate building height reconstruction from noisy, irregular TomoSAR point clouds by introducing a learning-based dual-topology network that alternates between a point-branch and a grid-branch to denoise and inpaint missing regions, producing high-resolution height maps. This first proof-of-concept demonstrates large-scale urban height mapping directly from TomoSAR data, with experiments on Munich and Berlin validating effectiveness and showing that incorporating optical imagery can further enhance results. The approach offers a weather-resilient, scalable route to 3D urban modeling, expanding TomoSAR utility beyond point clouds to continuous height fields. The work provides publicly available source code and lays groundwork for integrated, facade-aware urban reconstructions.
Abstract
Reliable building height estimation is essential for various urban applications. Spaceborne SAR tomography (TomoSAR) provides weather-independent, side-looking observations that capture facade-level structure, offering a promising alternative to conventional optical methods. However, TomoSAR point clouds often suffer from noise, anisotropic point distributions, and data voids on incoherent surfaces, all of which hinder accurate height reconstruction. To address these challenges, we introduce a learning-based framework for converting raw TomoSAR points into high-resolution building height maps. Our dual-topology network alternates between a point branch that models irregular scatterer features and a grid branch that enforces spatial consistency. By jointly processing these representations, the network denoises the input points and inpaints missing regions to produce continuous height estimates. To our knowledge, this is the first proof of concept for large-scale urban height mapping directly from TomoSAR point clouds. Extensive experiments on data from Munich and Berlin validate the effectiveness of our approach. Moreover, we demonstrate that our framework can be extended to incorporate optical satellite imagery, further enhancing reconstruction quality. The source code is available at https://github.com/zhu-xlab/tomosar2height.