PolyRoof: Precision Roof Polygonization in Urban Residential Building with Graph Neural Networks
Chaikal Amrullah, Daniel Panangian, Ksenia Bittner
TL;DR
The paper tackles the challenge of detailed roof polygonization in urban environments using aerial imagery. It extends Re:PolyWorld with PolyRoof by adding attention-based backbones and an area segmentation loss to handle complex residential roof geometries. On the RoofOpt dataset, the approach achieves improvements in point position accuracy ($RMSE$) and line distance accuracy, with a Reconstruction Score of $91.99\%$, demonstrating the value of advanced neural architectures for high-detail roof reconstruction. The work highlights practical implications for urban planning and building data generation, while acknowledging remaining gaps in edge completeness and the need for further generalization across geometric complexity.
Abstract
The growing demand for detailed building roof data has driven the development of automated extraction methods to overcome the inefficiencies of traditional approaches, particularly in handling complex variations in building geometries. Re:PolyWorld, which integrates point detection with graph neural networks, presents a promising solution for reconstructing high-detail building roof vector data. This study enhances Re:PolyWorld's performance on complex urban residential structures by incorporating attention-based backbones and additional area segmentation loss. Despite dataset limitations, our experiments demonstrated improvements in point position accuracy (1.33 pixels) and line distance accuracy (14.39 pixels), along with a notable increase in the reconstruction score to 91.99%. These findings highlight the potential of advanced neural network architectures in addressing the challenges of complex urban residential geometries.