Site-specific information concerning fuel hazard characteristics is needed to support wildfire management interventions and fuel hazard reduction programs. Currently, routine visual assessments provide subjective information, with the resulting estimate of fuel hazard varying due to observer experience and the rigor applied in making assessments. Terrestrial remote sensing techniques have been demonstrated to be capable of capturing quantitative information on the spatial distribution of biomass to inform fuel hazard assessments. This paper explores the use of image-based point clouds generated from imagery captured using a low-cost compact camera for describing the fuel hazard within the surface and near-surface layers. Terrestrial imagery was obtained at three distances for five target plots. Subsets of these images were then processed to determine the effect of varying overlap and distribution of image captures. The majority of the point clouds produced using this image-based technique provide an accurate representation of the 3D structure of the surface and near-surface fuels. Results indicate that high image overlap and pixel size are critical; multi-angle image capture is shown to be crucial in providing a representation of the vertical stratification of fuel. Terrestrial image-based point clouds represent a viable technique for low cost and rapid assessment of fuel structure.
Accurate three-dimensional representations of tree structure are essential for fire modeling, radiative transfer simulation, synthetic data generation, and digital twins of forests, yet detailed 3D structure is rarely available at required scales. Current approaches approximate crowns with smooth geometric primitives, discarding the clumping, gaps, and irregular branching present in real trees. We present TreeFlow, a conditional flow matching model that generates realistic 3D tree point clouds from species, acquisition platform, and height. The model uses a transformer trained on real laser scanning data from the FOR-species20K benchmark to learn a velocity field transporting samples from a Gaussian distribution to the source data distribution. We evaluate generation quality by comparing conditioning and distributional fidelity metrics to scans of real trees. Generated trees match or approach the intra-class baseline on five of six metrics, with a Chamfer distance of 0.581 m versus 0.559 m for real trees of the same genus and height class. Performance is strongest below 25 m and degrades with increasing height. TreeFlow is the first flow matching model to produce 3D tree point clouds from scalar inventory attributes using real laser scanning data.
Accurate three-dimensional representations of tree structure are essential for fire modeling, radiative transfer simulation, synthetic data generation, and digital twins of forests, yet detailed 3D structure is rarely available at required scales. Current approaches approximate crowns with smooth geometric primitives, discarding the clumping, gaps, and irregular branching present in real trees. We present TreeFlow, a conditional flow matching model that generates realistic 3D tree point clouds from species, acquisition platform, and height. The model uses a transformer trained on real laser scanning data from the FOR-species20K benchmark to learn a velocity field transporting samples from a Gaussian distribution to the source data distribution. We evaluate generation quality by comparing conditioning and distributional fidelity metrics to scans of real trees. Generated trees match or approach the intra-class baseline on five of six metrics, with a Chamfer distance of 0.581 m versus 0.559 m for real trees of the same genus and height class. Performance is strongest below 25 m and degrades with increasing height. TreeFlow generates individual-tree point clouds conditioned on scalar inventory attributes using a model trained entirely on real laser scanning data.