Quantifying the Accuracy of UAS-Lidar Individual Tree Detection Methods Across Height and Diameter at Breast Height Sizes in Complex Temperate Forests

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(6), P. 1010 - 1010

Published: March 13, 2025

Unpiloted aerial systems (UAS) and light detection ranging (lidar) sensors provide users with an increasingly accessible mechanism for precision forestry. As these technologies are further adopted, questions arise as to how select processing methods influencing subsequent high-resolution modelling analysis. This study addresses specific individual tree (ITD) impact the successful of trees varying sizes within complex forests. First, while many studies have compared ITD over several sites, algorithms, or sets parameters based on a singular validation metric, this quantifies 10 perform across tree-height size quartiles diameter at breast height (dbh) quartiles. In total, 1000 reference from 20 species three temperate forest sites were analyzed average point density 826.8 pts/m2. The results indicate that four classes, highest overall F-score (0.7344) was achieved F-scores 0.857 largest 0.633 smallest class. To expand analysis, generalized linear models used compare top performing worst method each variable site along continuous gradient. analysis suggests clear distinctions in performance (true positive false rates) method. UAS-lidar must ensure demonstrated validated ways communicate their relative effectiveness all sizes. Without such consideration, show surveys management conducted using may not accurately characterize present

Language: Английский

---

Three-Dimensional Reconstruction of Forest Scenes with Tree–Shrub–Grass Structure Using Airborne LiDAR Point Cloud

Forests, Journal Year: 2024, Volume and Issue: 15(9), P. 1627 - 1627

Published: Sept. 15, 2024

Fine three-dimensional (3D) reconstruction of real forest scenes can provide a reference for forestry digitization and resource management applications. Airborne LiDAR technology valuable data large-area scene reconstruction. This paper proposes 3D method complex with trees, shrubs, grass, based on airborne point clouds. First, vertical distribution characteristics are used to segment tree, shrub, ground–grass points from an cloud. For points, grid model is constructed. tree hierarchical canopy fitting proposed construct trunk model, crown constructed the α-shape algorithm. shrub directly Finally, models spatially combined achieve scenes. Taking six plots located in Hebei, Yunnan, Guangxi provinces China Baden-Württemberg Germany as study areas, experimental results show that accuracy individual segmentation reaches 87.32%, 60.00%, height grass evaluated RMSE < 0.15 m, volume assessed R2 > 0.848 0.904, respectively. Furthermore, we compared this simplified cloud voxel models. The demonstrate modeling approach meet demand high-accuracy lightweight

Language: Английский

---

A comprehensive review on tree detection methods using point cloud and aerial imagery from unmanned aerial vehicles

Computers and Electronics in Agriculture, Journal Year: 2024, Volume and Issue: 227, P. 109476 - 109476

Published: Oct. 1, 2024

Language: Английский

---

Accuracy Assessment of Advanced Laser Scanner Technologies for Forest Survey Based on Three-Dimensional Point Cloud Data

Sustainability, Journal Year: 2024, Volume and Issue: 16(23), P. 10636 - 10636

Published: Dec. 4, 2024

Forests play a crucial role in carbon sequestration and climate change mitigation, offering ecosystem services, biodiversity conservation, water resource management. As global efforts to reduce greenhouse gas emissions intensify, the demand for accurate spatial information monitor forest conditions assess absorption capacity has grown. LiDAR (Light Detection Ranging) emerged as transformative tool, providing high-resolution 3D data detailed analysis of attributes, including tree height, canopy structure, biomass distribution. Unlike traditional manpower-intensive surveys, which are time-consuming often limited accuracy, offers more efficient reliable solution. This study evaluates accuracy applicability advanced technologies—drone-mounted, terrestrial, mobile scanners—for generating data. The results show that terrestrial achieved highest precision diameter at breast height (DBH) measurements, with RMSE values 0.66 cm 0.91 m, respectively. Drone-mounted demonstrated excellent efficiency large-scale while offered portability speed but required further improvement (e.g., RMSE: DBH 0.76 cm, 1.83 m). By comparing these technologies, this identifies their strengths, limitations, optimal application scenarios, contributing management practices assessments.

Language: Английский

---

Evaluation of Tree Object Segmentation Performance for Individual Tree Recognition Using Remote Sensing Techniques Based on Urban Forest Green Structures

Land, Journal Year: 2024, Volume and Issue: 13(11), P. 1856 - 1856

Published: Nov. 7, 2024

This study evaluated whether tree object segmentation using remote sensing techniques could be effectively conducted according to the green structures of urban forests. The used were handheld LiDAR and UAV-based photogrammetry. data collected from both methods merged complement each other’s limitations. area classified into three types based on distance between canopy trees presence shrubs. ability individually classify within was then evaluated. evaluation method assess success rate by comparing actual number trees, which visually counted in field, with objects study. To perform semantic objects, a preprocessing step extract only related through techniques. steps included merging, noise removal, separation DTM DSM, areas structures. analysis results showed that recognition not efficient when complex mixed, highest present, canopies did overlap. Therefore, observing high-density areas, algorithm’s variables should adjusted narrow range, additional observations winter are needed compensate for obscured leaves. By improving collection systematizing structures, process can enhanced.

Language: Английский

---

Differences in detection results of trees of different species in the canopy of a different-aged polydominant broadleaved stand using the YOLO neural network

Doklady Meždunarodnoj konferencii "Matematičeskaâ biologiâ i bioinformatika", Journal Year: 2024, Volume and Issue: 10

Published: Oct. 27, 2024

Language: Английский

---