3DFin: a software for automated 3D forest inventories from terrestrial point clouds

Abstract

Abstract Accurate and efficient forest inventories are essential for effective forest management and conservation. The advent of ground-based remote sensing has revolutionized the data acquisition process, enabling detailed and precise 3D measurements of forested areas. Several algorithms and methods have been developed in the last years to automatically derive tree metrics from such terrestrial/ground-based point clouds. However, few attempts have been made to make these automatic tree metrics algorithms accessible to wider audiences by producing software solutions that implement these methods. To fill this major gap, we have developed 3DFin, a novel free software program designed for user-friendly, automatic forest inventories using ground-based point clouds. 3DFin empowers users to automatically compute key forest inventory parameters, including tree Total Height, Diameter at Breast Height (DBH), and tree location. To enhance its user-friendliness, the program is open-access, cross-platform, and available as a plugin in CloudCompare and QGIS as well as a standalone in Windows. 3DFin capabilities have been tested with Terrestrial Laser Scanning, Mobile Laser Scanning, and terrestrial photogrammetric point clouds from public repositories across different forest conditions, achieving nearly full completeness and correctness in tree mapping and highly accurate DBH estimations (root mean squared error <2 cm, bias <1 cm) in most scenarios. In these tests, 3DFin demonstrated remarkable efficiency, with processing times ranging from 2 to 7 min per plot. The software is freely available at: https://github.com/3DFin/3DFin.

Plant phenotyping relevance

森林個体の樹高・胸高直径などの植物形質を点群から自動抽出するソフトウェアの開発と技術検証が中心であり、植物フェノタイピング手法に該当する。

Code and data availability

The paper's DBH/tree-metric analysis was run on the public SilviLaser 2021 Benchmark Dataset (TU Wien Research Data, DOI 10.48436/afdjq-ce434), and the authors' analysis software 3DFin is publicly available (GitHub releases, CloudCompare plugin, PyPI). Both are paper-specific, public, and actionable.