Thermoxels: a voxel-based method to generate simulation-ready 3D thermal models

Abstract In the European Union, buildings account for 42% of energy use and 35% of greenhouse gas emissions. Since most buildings will still be in use by 2050, retrofitting is crucial for reducing emissions. Current building assessment methods mainly rely on qualitative thermal imaging, which limits data-driven decisions for energy savings. On the other hand, quantitative assessments using finite element analysis (FEA) provide precise insights but require manual CAD design by experts. Recent advances in 3D reconstruction, such as Neural Radiance Fields and Gaussian Splatting, enable precise 3D modeling from sparse images but lack the defined volumes and interfaces needed for FEA. We propose Thermoxels, a novel voxel-based method able to generate FEA-compatible models, including both geometry and temperature, from a RGB and thermal images. Using pairs of RGB and thermal images as input, Thermoxels represents a scene’s geometry as a set of voxels comprising color and temperature information. The color and temperature parameters are optimized via backpropagation against ground truth values. Then, voxels are filtered based on a density threshold, converting the scene representation into a tetrahedral mesh suitable for FEA. We demonstrate Thermoxels’ capability to generate FEA-compatible RGB+Thermal meshes of 3D scenes, a task where current state-of-the-art fall short. We conduct heat conduction simulations using FEA and Thermoxels’ thermal reconstruction as initial state across five different scene, demonstrating the convergence of these simulation. Additionally, we evaluate Thermoxels’ image synthesis capabilities, achieving an average thermal MAE of 2.77±2.76,◦C, and discuss the limitations of existing metrics in evaluating mesh quality.