講演詳細

防災・水文部門[Forest Disaster Prevention and Hydrology]

日付	2026年3月18日
開始時刻	ポスター発表
会場名	多目的ホール
講演番号	PJ-57
発表題目	DBH Estimation from Mobile LiDAR in Natural Temperate Mixed Forests DBH Estimation from Mobile LiDAR in Natural Temperate Mixed Forests
所属	九州大学
要旨本文	DBH is fundamental to forest inventory and biomass estimation, but conventional field measurement is labor intensive. Mobile LiDAR can reduce effort, yet DBH estimation in mixed forests is challenging because leaning or curved stems, buttresses, and multi-stemming complicate breast-height geometry. We present an open-source Python workflow that tracks stems across multiple height slices, reconstructs 3D stem axes, and estimates DBH from axis-orthogonal sections, with FAO-consistent height adjustment when irregular bases are detected. Compared with existing open-source DBH software, it shows more complete stem identification and stem-geometry labeling. Against field DBH inventory, it reduces error relative to a published baseline (RMSE: 1.18 cm vs. 2.01 cm). Future work will scale processing to the full 1 ha plot and conduct stem-form-specific error analyses to diagnose dominant error sources and further improve robustness.
著者氏名	○PARK, JI HYEOK1 ・ Gomez, Christopher2 ・ Kume, Tomonori1
著者所属	1Kyushu University ・ 2Kobe University
キーワード	1ha Scale Forest Inventory, Algorithm, Point Cloud Data
Key word	1ha Scale Forest Inventory, Algorithm, Point Cloud Data