Delaunay Canopy: Building Wireframe Reconstruction from Airborne LiDAR Point Clouds via Delaunay Graph

2026-04-02

AI Summary

This work addresses the challenge of inaccurate architectural wireframe reconstruction in noisy, sparse point clouds or regions containing interior corners. It introduces, for the first time, the use of Delaunay graphs as geometric priors to construct an adaptive search space. By scoring graph elements, the method reconstructs geometric manifolds and estimates regional curvature, which jointly guide the precise selection of corners and line segments. This approach enables adaptive modeling of complex architectural geometries and significantly improves reconstruction robustness and accuracy in challenging regions. Evaluations on the Building3D Tallinn dataset and entry-level benchmarks demonstrate state-of-the-art performance.

Abstract

Reconstructing building wireframe from airborne LiDAR point clouds yields a compact, topology-centric representation that enables structural understanding beyond dense meshes. Yet a key limitation persists: conventional methods have failed to achieve accurate wireframe reconstruction in regions afflicted by significant noise, sparsity, or internal corners. This failure stems from the inability to establish an adaptive search space to effectively leverage the rich 3D geometry of large, sparse building point clouds. In this work, we address this challenge with Delaunay Canopy, which utilizes the Delaunay graph as a geometric prior to define a geometrically adaptive search space. Central to our approach is Delaunay Graph Scoring, which not only reconstructs the underlying geometric manifold but also yields region-wise curvature signatures to robustly guide the reconstruction. Built on this foundation, our corner and wire selection modules leverage the Delaunay-induced prior to focus on highly probable elements, thereby shaping the search space and enabling accurate prediction even in previously intractable regions. Extensive experiments on the Building3D Tallinn city and entry-level datasets demonstrate state-of-the-art wireframe reconstruction, delivering accurate predictions across diverse and complex building geometries.

Problem

• building wireframe reconstruction
• airborne LiDAR point clouds
• noise and sparsity
• internal corners
• adaptive search space

Innovation

• Delaunay graph
• wireframe reconstruction
• LiDAR point clouds
• geometric prior
• adaptive search space