A Unified Graph-based Framework for Scalable 3D Tree Reconstruction and Non-Destructive Biomass Estimation from Point Clouds Journal: Arxiv
Di Wang, Shi Li · 2025
Read A Unified Graph-based Framework for Scalable 3D Tree Reconstruction and Non-Destructive Biomass Estimation from Point Clouds Journal: Arxiv by Di Wang, Shi Li (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-15577.txt.
Introduction
Di Wang and Shi Li present in 2025 a graph-based method to rebuild trees in 3D from laser point clouds. Note that the text's own title differs from the catalog title, and the text governs here. Their goal is to estimate wood mass above ground without first stripping leaves.
Leaves usually force a separate sorting step that adds error. Instead, the method counts how many root paths pass through each point on the shortest-path tree. That count maps the branch structure in one run. Tests on ninety leafy trees, in tropics and temperate zones, beat a leading leaf-sorting pipeline.
Catalog record
| CREATOR | Di Wang, Shi Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-15577
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Unified Graph-based Framework for Scalable 3D Tree Reconstruction and Non-Destructive Biomass Estimation from Point Clouds Journal: Arxiv Di Wang, Shi Li Abstract Estimating forest above-ground biomass (AGB) is crucial for assessing carbon storage and supporting sustainable forest management. Quantitative Structural Model (QSM) offers a non-destructive approach to AGB estimation through 3D tree structural reconstruction. However, current QSM methods face significant limitations, as they are primarily designed for individual trees, depend on high-quality point cloud data from terrestrial laser scanning (TLS), and also require multiple pre-processing steps that hinder scalability and practical deployment. This study presents a novel unified framework that enables end-to-end processing of large-scale point clouds using an innovative graph-based pipeline. The proposed approach seamlessly integrates tree segmentation, leaf-wood separation and 3D skeletal reconstruction through dedicated graph operations including pathing and abstracting for tree topology reasoning. Comprehensive validation was conducted on datasets with varying leaf conditions (leaf-on and leaf-off), spatial scales (tree- and plot-level), and data sources (TLS and UAV-based laser scanning, ULS). Experimental results demonstrate strong performance under challenging conditions, particularly in leaf-on scenarios ( ∼ \sim 20% relative error) and low-density ULS datasets with partial coverage ( ∼ \sim 30% relative error). These findings indicate that the proposed framework provides a robust and scalable solution for large-scale, non-destructive AGB estimation. It significantly reduces dependency on specialized pre-processing tools and establishes ULS as a viable alternative to TLS. To our knowledge, this is the first method capable of enabling seamless, end-to-end 3D tree reconstruction at operational scales. This advancement substantially improves the feasibility of QSM-based AGB estimation, paving the way for broader applications in forest inventory and climate change research. 1 Introduction Trees play a crucial role in carbon sequestration, climate regulation, and biodiversity maintenance ( Bonan 2008 ) , serving as primary carriers of carbon stocks in terrestrial ecosystems. Accurate estimation of Above-Ground Biomass (AGB) over large spatial extents is essential for evaluating forest carbon stocks and ecosystem service functions ( Goetz et al. 2009 ) . In particular, precise AGB assessments are not only vital for ecological research but also fundamental to the implementation and verification of global climate policies, such as the Paris Agreement and the Reducing Emissions from Deforestation and Forest Degradation (REDD+) program ( Duncanson et al. 2019a ) . These initiatives rely on reliable carbon stock estimates to guide mitigation efforts and support carbon trading schemes. The direct method of measuring AGB involves harvesting and weighing trees. However, this approach is not only harmful to the trees but also labor-intensive, time-consuming, and requires substantial manual effort along with strict quality control ( Hyyppa et al. 2001 ) . Consequently, there is a growing need for more efficient and scalable methods that enable accurate AGB estimation without extensive fieldwork and in a manner that minimizes damage to the trees. Remote sensing techniques have developed rapidly over the past decades, providing data with higher spatial resolution, finer spectral channels, and various modalities. These advancements make remote sensing increasingly useful for large-scale AGB estimations ( Ometto et al. 2023 ) . Nonetheless, remote sensing methods still require accurate ground data for developing, calibrating, and validating their biomass models and products ( Labrière et al. 2023 ) . In general, ground reference data typically rely on allometric models ( Roxburgh et al. 2015 ) . However, these models often necessitate calibration through destructive measures. Many allometric models are either inaccurate or unsuitable for specific geographical regions, constituting a major source of error in remote sensing AGB estimates ( Chen et al. 2015 ) . Advances in Light Detection and Ranging (LiDAR) technology have provided a solid data foundation for tree structure analyses and biomass estimation ( Calders et al. 2020 ) . Due to its strong penetration capability and high spatial resolution, LiDAR is particularly effective at capturing the three-dimensional (3D) structural characteristics of trees, even in complex understory environments ( Liang et al. 2016 ) . A variety of LiDAR systems have been developed to address different forest mapping requirements. Terrestrial Laser Scanning (TLS) provides high-resolution, ground-based measurements that are ideal for detailed single-tree level structural analysis, while Airborne Laser Scanning (ALS) enables large-scale forest mapping, albeit with relatively lower point density. Mobile Laser Scanning (MLS) and UAV-based Laser Scanning (ULS) offer flexible deployment options for challenging terrains. Building upon these LiDAR-derived data sources, point cloud-based techniques have made significant progress across various aspects of tree structural analysis. These include the estimation of key tree parameters such as height, diameter at breast height (DBH), and crown dimensions ( Wulder et al. 2012 ) . A particularly useful technique developed over the past decade is 3D tree reconstruction from point cloud data. The so-called tree Quantitative Structural Model (QSM) enables tree branching structure reconstruction and volume estimation in a non-destructive manner ( Demol et al. 2022 ) . Essentially, QSM is a 3D geometric and topological representation of a tree that captures its branching architecture with measurable attributes such as branch lengths, diameters, volumes, and spatial orientations ( Ali et al. 2025 ) . When combined with information on tree wood density, the QSM-estimated volume can be converted into AGB or even carbon storage, supporting a wide range of ecological and environmental studies ( Calders et al. 2015 ) . This approach provides an exciting and promising way to accurately estimate tree AGB without destructive measures. Numerous studies have demonstrated its superior performance compared to conventional allometric models, which require careful calibration ( Calders et al. 2015 ; Momo Takoudjou et al. 2018 ; Gonzalez de Tanago et al. 2018 ) . However, despite significant progress in QSM techniques, several challenges continue to hinder its application in large-scale forest assessments. QSM typically operates on individual trees and requires single-tree point clouds as input. In practice, two prerequisite steps, individual tree segmentation and leaf-wood separation, must be performed prior to applying QSM at scale, and each of these steps can be technically challenging ( Martin-Ducup et al. 2021b ) . Second, current QSM methods generally require high-density and well-covered point cloud data. As a result, multi-scan TLS data are most commonly used ( Demol et al. 2022 ) . However, acquiring large-scale TLS data can be labor-intensive and time-consuming, particularly in complex or difficult-to-access terrain ( Ye et al. 2019 ) . Lastly, many existing approaches rely on complex, manually tuned workflows that are not easily transferable across diverse or multi-source datasets ( Zhang et al. 2023 ) . Collectively, these limitations constrain the scalability and practical applicability of QSM-based biomass estimation in real-world forest monitoring scenarios. In this study, we propose a unified 3D tree reconstruction framework based on graph processing, designed to enable scalable and fully automated 3D tree reconstruction from large-scale point cloud data, with a primary focus on its application to non-destructive AGB estimation. The core idea involves transforming point cloud data into a graph-based representation, followed by a series of graph operations to perform key tasks such as individual tree segmentation, leaf-wood separation, and 3D skeletal reconstruction. This novel framework demonstrates strong robustness to variations in point cloud density and foliage conditions, achieving reliable performance even on sparse ULS data. We evaluate the effectiveness of the framework across various data acquisition