PolyGNN: Polyhedron-based Graph Neural Network for 3D Building Reconstruction from Point Clouds
This work addresses the problem of creating detailed 3D building models from point clouds for urban planning or simulation, representing an incremental improvement through a novel method for a known bottleneck in handling arbitrary-shaped polyhedra.
The authors tackled 3D building reconstruction from point clouds by proposing PolyGNN, a polyhedron-based graph neural network that assembles primitives via graph node classification, achieving watertight and compact reconstructions with demonstrated efficiency for large-scale applications.
We present PolyGNN, a polyhedron-based graph neural network for 3D building reconstruction from point clouds. PolyGNN learns to assemble primitives obtained by polyhedral decomposition via graph node classification, achieving a watertight and compact reconstruction. To effectively represent arbitrary-shaped polyhedra in the neural network, we propose a skeleton-based sampling strategy to generate polyhedron-wise queries. These queries are then incorporated with inter-polyhedron adjacency to enhance the classification. PolyGNN is end-to-end optimizable and is designed to accommodate variable-size input points, polyhedra, and queries with an index-driven batching technique. To address the abstraction gap between existing city-building models and the underlying instances, and provide a fair evaluation of the proposed method, we develop our method on a large-scale synthetic dataset with well-defined ground truths of polyhedral labels. We further conduct a transferability analysis across cities and on real-world point clouds. Both qualitative and quantitative results demonstrate the effectiveness of our method, particularly its efficiency for large-scale reconstructions. The source code and data are available at https://github.com/chenzhaiyu/polygnn.