pith. sign in

arxiv: 2605.02201 · v2 · submitted 2026-05-04 · 💻 cs.CV

Super-Resolution of Airborne Laser Scanning Point Clouds for Forest Inventory

Jinyuan Shao , Sangyoong Park , Chunxi Zhao , Ayman Habib , Songlin Fei This is my paper

Pith reviewed 2026-05-11 00:59 UTC · model grok-4.3

classification 💻 cs.CV
keywords point cloud super-resolutionairborne laser scanningforest inventorystem detectiondeep learningU-NetLiDARtree reconstruction
0
0 comments X

The pith

A voxel-based U-Net model super-resolves sparse airborne laser point clouds to enable accurate stem detection and diameter estimation in forests.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces 3DFSR, a voxel-based convolutional neural network with U-Net architecture, designed to increase the density and reduce noise in airborne laser scanning point clouds of forests. A sympathetic reader would care because this could allow wide-area forest inventories using affordable airborne data to achieve accuracy levels previously requiring expensive ground-based scanning. The model was tested on temperate forests in the U.S. and boreal forests in Germany, showing that the enhanced clouds enable direct use of stem detection and reconstruction methods developed for denser data. It reports specific improvements in detection F1 scores, diameter estimation errors, and reconstruction fidelity to reference measurements.

Core claim

The 3DFSR model generates finer point clouds of tree structures from ALS data, achieving 0.249 m Chamfer Distance and 2.711 m Hausdorff Distance, which in turn allows stem detection algorithms to reach an F1 score of 0.97, DBH estimation with 6.43 cm RMSE via circle fitting, and stem reconstructions with 0.170 m Chamfer Distance and 0.95 R-squared volume correlation to MLS references.

What carries the argument

3DFSR, a voxel-based CNN with U-Net architecture that learns to map low-density ALS forest point clouds to higher-density, lower-noise versions.

If this is right

  • Stem detection F1 score rises from 0.71 on raw ALS to 0.97 on 3DFSR outputs.
  • DBH root mean square error falls from 13.45 cm using allometric equations to 6.43 cm using circle fitting on enhanced clouds.
  • Stem reconstructions from 3DFSR match MLS references with 0.170 m Chamfer distance, 0.377 m Hausdorff distance, and 0.95 R2 for volume.
  • The approach works across input densities from 10 to 1700 points per square meter.
  • It generalizes to different LiDAR platforms without needing transfer learning or retraining.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Forestry agencies could retrofit existing ALS datasets for high-resolution inventory without new acquisitions.
  • Similar voxel-based super-resolution networks might adapt to other 3D remote sensing tasks such as urban canopy modeling.
  • Verification on additional forest types would test whether the learned super-resolution preserves geometry universally.
  • Integration with downstream machine learning for species classification or biomass estimation could follow from the cleaner inputs.

Load-bearing premise

Super-resolved points must accurately represent actual tree geometry and avoid introducing artifacts that bias stem detection, circle fitting, or volume reconstruction.

What would settle it

Running stem detection and DBH circle fitting on 3DFSR point clouds from a new forest site and finding large discrepancies with independent field measurements of the same trees would falsify the effectiveness claim.

Figures

Figures reproduced from arXiv: 2605.02201 by Ayman Habib, Chunxi Zhao, Jinyuan Shao, Sangyoong Park, Songlin Fei.

Figure 1
Figure 1. Figure 1: Examples of ALS and TLS/MLS LiDAR point clouds used in this study. 3.2. Data collection We introduce data collection of low-resolution ALS data and high-resolution TLS/MLS data used in this study view at source ↗
Figure 1
Figure 1. Figure 1: Examples of ALS and TLS/MLS LiDAR point clouds used in this study. 3.2. Data collection We introduce data collection of low-resolution ALS data and high-resolution TLS/MLS data used in this study [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Octree-based voxelization. Conv 1x1x1 Group Norm ReLU Conv 3x3x3 Group Norm ReLU Conv 1x1x1 Group Norm ReLU Addition Octree based residual block Skip connection Input octree Output octree Ls Lr view at source ↗
Figure 2
Figure 2. Figure 2: Octree-based voxelization. Conv 1x1x1 Group Norm ReLU Conv 3x3x3 Group Norm ReLU Conv 1x1x1 Group Norm ReLU Addition Octree based residual block Skip connection Input octree Output octree Ls Lr [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Model architecture of 3DFSR. features into a higher-dimensional latent space. Each level then applies two residual residual blocks, where each resid￾ual block consists of three consecutive octree convolutions with group normalization and ReLU activation, plus a skip connection that adds the block input to its output. This design helps stabilize training and preserves fine-grained geometry features. Between… view at source ↗
Figure 3
Figure 3. Figure 3: Model architecture of 3DFSR. features into a higher-dimensional latent space. Each level then applies two residual residual blocks, where each resid￾ual block consists of three consecutive octree convolutions with group normalization and ReLU activation, plus a skip connection that adds the block input to its output. This design helps stabilize training and preserves fine-grained geometry features. Between… view at source ↗
Figure 4
Figure 4. Figure 4: Super-resolution results obtained from naive models of PU-Net and GradPU in five plots view at source ↗
Figure 4
Figure 4. Figure 4: Super-resolution results obtained from naive models of PU-Net and GradPU in five plots [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Super-resolution results with different methods in five plots. Negative (FN), completeness, omission, commission and F1 score. 5.2.2. Stem detection results We present stem mapping results obtained using the ALS-based DFT toolbox on both the original ALS point clouds and the super-resolution point clouds in Tab. 4, and results of TLS/MLS-based method, LeSSO, in Tab. 5. Since Shao et al.: Preprint submitted… view at source ↗
Figure 5
Figure 5. Figure 5: Super-resolution results with different methods in five plots. Negative (FN), completeness, omission, commission and F1 score. 5.2.2. Stem detection results We present stem mapping results obtained using the ALS-based DFT toolbox on both the original ALS point clouds and the super-resolution point clouds in Tab. 4, and results of TLS/MLS-based method, LeSSO, in Tab. 5. Since Shao et al.: Preprint submitted… view at source ↗
Figure 6
Figure 6. Figure 6: Super-resoltuion results with different methods on single trees of dominant species in five plots. LeSSO is designed for high-density point clouds, it is only applied to the super-resolution data. For the results obtained with the DFT toolbox, the over￾all performance shows marginal changes when moving from the original ALS data to the super resolution point clouds, with F1 score from 0.71 to 0.70. In term… view at source ↗
Figure 6
Figure 6. Figure 6: Super-resoltuion results with different methods on single trees of dominant species in five plots. LeSSO is designed for high-density point clouds, it is only applied to the super-resolution data. For the results obtained with the DFT toolbox, the over￾all performance shows marginal changes when moving from the original ALS data to the super resolution point clouds, with F1 score from 0.71 to 0.70. In term… view at source ↗
Figure 7
Figure 7. Figure 7: DBH estimation with original ALS point clouds using allometric equations view at source ↗
Figure 7
Figure 7. Figure 7: DBH estimation with original ALS point clouds using allometric equations [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: DBH estimation with super-resolution ALS point clouds using circle fitting. from 10.33 to 17.33 cm view at source ↗
Figure 8
Figure 8. Figure 8: DBH estimation with super-resolution ALS point clouds using circle fitting. from 10.33 to 17.33 cm [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Stem reconstruction based on super-resolution (SR) point clouds and Mobile Laser Scanning (MLS) point clouds. Plot I Plot II Plot III view at source ↗
Figure 9
Figure 9. Figure 9: Stem reconstruction based on super-resolution (SR) point clouds and Mobile Laser Scanning (MLS) point clouds. Plot I Plot II Plot III [PITH_FULL_IMAGE:figures/full_fig_p013_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Stem volume comparison between stem reconstructed from super-resolution point clouds and MLS point clouds. densities and use them as input separately. Downsampled point densities are set to 10, 25, 50, 75, 100, 250, 500, 1000, 1500 and 1700. This range of point densities covers from the linear-model ALS data (10 points/m2 ) to high￾resolution UAV LiDAR data (1700 points/m2 ). For all mod￾els in this exper… view at source ↗
Figure 10
Figure 10. Figure 10: Stem volume comparison between stem reconstructed from super-resolution point clouds and MLS point clouds. densities and use them as input separately. Downsampled point densities are set to 10, 25, 50, 75, 100, 250, 500, 1000, 1500 and 1700. This range of point densities covers from the linear-model ALS data (10 points/m2 ) to high￾resolution UAV LiDAR data (1700 points/m2 ). For all mod￾els in this exper… view at source ↗
Figure 11
Figure 11. Figure 11: Super-resolution results using different densities as the input view at source ↗
Figure 11
Figure 11. Figure 11: Super-resolution results using different densities as the input [PITH_FULL_IMAGE:figures/full_fig_p014_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Performance variation on Chamfer Distance (left) and Hausdoff Distance (right) based on the change of input point densities. produce super-resolution point clouds for all input densities. All generated super-resolution point clouds have dense and clean stem points, but have more clear canopy structures as the input densities increased. For some crowded areas, there are dense points between stems, and this… view at source ↗
Figure 12
Figure 12. Figure 12: Performance variation on Chamfer Distance (left) and Hausdoff Distance (right) based on the change of input point densities. produce super-resolution point clouds for all input densities. All generated super-resolution point clouds have dense and clean stem points, but have more clear canopy structures as the input densities increased. For some crowded areas, there are dense points between stems, and this… view at source ↗
Figure 13
Figure 13. Figure 13: Performance variation on Chamfer Distance (left) and Hausdorff Distance (right) based on the change of input point densities view at source ↗
Figure 13
Figure 13. Figure 13: Performance variation on Chamfer Distance (left) and Hausdorff Distance (right) based on the change of input point densities [PITH_FULL_IMAGE:figures/full_fig_p015_13.png] view at source ↗
read the original abstract

Airborne Laser Scanning (ALS) can collect point clouds across large areas, enabling large-scale forest inventory. However, ALS point clouds are sparse and noisy, resulting in inaccurate individual-tree-level forest inventory, such as stem localization and tree size estimation. To overcome this problem, we propose a deep learning model, 3D Forest Super Resolution (3DFSR), to simultaneously improve point density and reduce noise for ALS forest point cloud. 3DFSR is a voxel-based CNN with a U-Net architecture. The proposed 3DFSR is evaluated on ALS point clouds collected in both temperate forests in the U.S. and boreal forests in Germany. Experimental results demonstrate that 3DFSR can generate finer point clouds of tree structure than other state-of-the-art point cloud super-resolution algorithms, achieving 0.249 m Chamfer Distance and 2.711 m Hausdorff Distance. Furthermore, to verify the effectiveness of 3DFSR point clouds in forest inventory, we conduct stem detection, DBH measurements, and stem reconstruction on both original ALS point clouds and 3DFSR enhanced point clouds. We find that stem detection and reconstruction algorithms developed for TLS/MLS point clouds can directly work on our 3DFSR point clouds, and DBH can be derived with circle-fitting method. F1 score of stem detection is improved from 0.71 on original ALS point clouds to 0.97 on 3DFSR point clouds; DBH estimation improves from 13.45 cm RMSE using allometric equations to 6.43 cm using circle fitting; comparing to stems reconstruction from MLS point clouds, stem reconstructed from 3DFSR point clouds has 0.170 m of Chamfer Distance and 0.377 m of Hausdorff Distance, and 0.95 R2 volume estimation. Finally, we find that the proposed 3DFSR is applicable to process point densities from 10 to 1700 points/m2; it also can be generalized across data collected from different LiDAR platforms without transfer learning.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 1 minor

Summary. The paper proposes 3DFSR, a voxel-based U-Net CNN for super-resolving sparse, noisy ALS forest point clouds to enable direct use of TLS/MLS stem detection, DBH circle-fitting, and reconstruction algorithms. On US temperate and German boreal datasets it reports Chamfer Distance 0.249 m / Hausdorff 2.711 m for the super-resolved clouds, stem-detection F1 rising from 0.71 to 0.97, DBH RMSE falling from 13.45 cm to 6.43 cm, and stem reconstruction Chamfer 0.170 m / volume R² 0.95 versus MLS references; it further claims applicability across 10–1700 pts/m² densities and cross-platform generalization without transfer learning.

Significance. If the geometric fidelity claims hold, the work would be significant for scaling individual-tree inventory over large areas by allowing existing high-resolution algorithms to operate on routinely collected ALS data. Concrete gains on two distinct forest types and three downstream tasks, plus the density-range and cross-platform statements, provide a practical bridge between ALS and TLS/MLS pipelines.

major comments (3)
  1. [Abstract and Results] Abstract / Results: the central performance claims (F1 0.71→0.97, DBH RMSE 13.45→6.43 cm, volume R² 0.95) rest on downstream metrics that can improve even when added points contain local biases or smoothing artifacts; the manuscript provides no local error maps, stem-axis deviation statistics, or qualitative branch-level comparisons that would confirm the super-resolved geometry remains within the tolerance of circle-fitting and reconstruction pipelines.
  2. [Methods] Methods / Experimental setup: no information is given on training/validation/test splits, voxelization grid size, loss weighting, or ablation studies that isolate the contribution of the U-Net super-resolution step; without these details the reported Chamfer/Hausdorff numbers and generalization statements (10–1700 pts/m², cross-platform) cannot be reproduced or stress-tested.
  3. [Evaluation] Evaluation: the two-site (temperate US + boreal Germany) test does not include cross-sensor or cross-species hold-out experiments; the claim of platform-independent generalization therefore lacks the concrete test required to support deployment beyond the two sensor/forest regimes examined.
minor comments (1)
  1. [Abstract] Abstract contains repeated phrasing (“we find that”) that could be tightened for clarity.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the positive assessment of the work's significance and for the detailed major comments. We respond to each point below and outline the revisions we plan to make to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and Results] Abstract / Results: the central performance claims (F1 0.71→0.97, DBH RMSE 13.45→6.43 cm, volume R² 0.95) rest on downstream metrics that can improve even when added points contain local biases or smoothing artifacts; the manuscript provides no local error maps, stem-axis deviation statistics, or qualitative branch-level comparisons that would confirm the super-resolved geometry remains within the tolerance of circle-fitting and reconstruction pipelines.

    Authors: We recognize the validity of this concern; downstream task performance alone does not guarantee that the super-resolved points are free of local biases that could affect circle fitting or reconstruction. While the reported Chamfer and Hausdorff distances offer global measures of fidelity, we agree that additional local analyses would be beneficial. In the revised manuscript, we will add local error maps visualizing point deviations from reference MLS data, compute stem-axis deviation statistics for the detected stems, and include qualitative side-by-side comparisons of branch-level structures in the super-resolved ALS, original ALS, and MLS point clouds. revision: yes

  2. Referee: [Methods] Methods / Experimental setup: no information is given on training/validation/test splits, voxelization grid size, loss weighting, or ablation studies that isolate the contribution of the U-Net super-resolution step; without these details the reported Chamfer/Hausdorff numbers and generalization statements (10–1700 pts/m², cross-platform) cannot be reproduced or stress-tested.

    Authors: We agree that these methodological details are essential for reproducibility and were omitted from the original submission. We will revise the Methods section to include: (i) a description of the training/validation/test splits, which were conducted at the plot level to prevent leakage; (ii) the specific voxelization grid size employed; (iii) the formulation and weighting of the loss function; and (iv) ablation studies that quantify the impact of the U-Net architecture and loss components on the super-resolution performance and downstream metrics. revision: yes

  3. Referee: [Evaluation] Evaluation: the two-site (temperate US + boreal Germany) test does not include cross-sensor or cross-species hold-out experiments; the claim of platform-independent generalization therefore lacks the concrete test required to support deployment beyond the two sensor/forest regimes examined.

    Authors: The evaluation already spans two distinct forest types (temperate US and boreal Germany) and different ALS platforms, with the model applied without transfer learning, supporting the generalization claim to some extent. Nevertheless, we concur that dedicated hold-out experiments (e.g., training exclusively on one dataset and testing on the other) would provide more rigorous evidence. We will either perform and report such experiments in the revision or, if data constraints prevent it, explicitly discuss the scope and limitations of the current generalization results in the manuscript. revision: partial

Circularity Check

0 steps flagged

No significant circularity; empirical validation is independent of internal definitions

full rationale

The paper proposes a voxel-based U-Net (3DFSR) for ALS point cloud super-resolution and evaluates it on independent reference data from temperate US and boreal German forests. All reported metrics (Chamfer/Hausdorff distances, stem detection F1, DBH RMSE, volume R²) are computed directly against external MLS/TLS ground truth. No equations, predictions, or claims reduce by construction to fitted parameters, self-defined quantities, or self-citation chains; the derivation chain consists of standard supervised training followed by out-of-sample empirical testing across sites and densities. This is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The central claims rest on empirical results from a supervised deep learning model trained on forest point cloud pairs; no additional free parameters, axioms, or invented entities beyond standard CNN training assumptions are introduced.

pith-pipeline@v0.9.0 · 5690 in / 1283 out tokens · 53426 ms · 2026-05-11T00:59:43.401923+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

113 extracted references · 113 canonical work pages

  1. [1]

    , title =

    Fortunato, S. , title =. Phys. Rep.-Rev. Sec. Phys. Lett. , volume =. 2010 , pages =

    work page 2010

  2. [2]

    Newman, M. E. J. and Girvan, M. , title =. Phys. Rev. E. , volume =. 2004 , pages =

    work page 2004

  3. [3]

    and Reinhardt, T

    Vehlow, C. and Reinhardt, T. and Weiskopf, D. , title =. IEEE Trans. Vis. Comput. Graph. , volume =. 2013 , pages =

    work page 2013

  4. [4]

    and Albert, R

    Raghavan, U. and Albert, R. and Kumara, S. , title =. Phys. Rev E. , volume =. 2007 , pages =

    work page 2007

  5. [5]

    2011 , pages =

    Robust network community detection using balanced propagation , journal =. 2011 , pages =

    work page 2011

  6. [6]

    and Li, S

    Lou, H. and Li, S. and Zhao, Y. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  7. [7]

    and Newman, M

    Clauset, A. and Newman, M. E. J. and Moore, C. , title =. Phys. Rev. E. , volume =. 2004 , pages =

    work page 2004

  8. [8]

    Blondel, V. D. and Guillaume, J. L. and Lambiotte, R. and Lefebvre, E. , title =. J. Stat. Mech.-Theory Exp. , volume =. 2008 , pages =

    work page 2008

  9. [9]

    and Campari, R

    Sobolevsky, S. and Campari, R. , title =. Phys. Rev. E. , volume =. 2014 , pages =

    work page 2014

  10. [10]

    and Barthelemy, M

    Fortunato, S. and Barthelemy, M. , title =. Proc. Natl. Acad. Sci. U. S. A. , volume =. 2007 , pages =

    work page 2007

  11. [11]

    2011 , pages =

    Unfolding communities in large complex networks: Combining defensive and offensive label propagation for core extraction , journal =. 2011 , pages =

    work page 2011

  12. [12]

    and Li, J

    Wang, X. and Li, J. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  13. [13]

    and Wang, X

    Li, J. and Wang, X. and Eustace, J. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  14. [14]

    Fabio, D. R. and Fabio, D. and Carlo, P. , title =. Sci. Rep. , volume =. 2013 , pages =

    work page 2013

  15. [15]

    and Wu, T

    Chen, Q. and Wu, T. T. and Fang, M. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  16. [16]

    and Wang, R

    Zhang, S. and Wang, R. and Zhang, X. , title =. Physica A. , volume =. 2007 , pages =

    work page 2007

  17. [17]

    and Petr\'oczi, A

    Nepusz, T. and Petr\'oczi, A. and N\'egyessy, L. and Bazs\'o, F. , title =. Phys. Rev. E. , volume =. 2008 , pages =

    work page 2008

  18. [18]

    and Liang, Z

    Fabricio, B. and Liang, Z. , title =. Soft Comput. , volume =. 2013 , pages =

    work page 2013

  19. [19]

    and Gao, L

    Sun, P. and Gao, L. and Han, S. , title =. Inf. Sci. , volume =. 2011 , pages =

    work page 2011

  20. [20]

    and Liu, D

    Wang, W. and Liu, D. and Liu, X. and Pan, L. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  21. [21]

    and Roberts, S

    Psorakis, I. and Roberts, S. and Ebden, M. and Sheldon, B. , title =. Phys. Rev. E. , volume =. 2011 , pages =

    work page 2011

  22. [22]

    and Yeung, D

    Zhang, Y. and Yeung, D. , title =. In Proc. ACM SIGKDD Conf. , year =

    work page

  23. [23]

    , title =

    Liu, J. , title =. Eur. Phys. J. B. , volume =. 2010 , pages =

    work page 2010

  24. [24]

    Havens, T. C. and Bezdek, J. C. and Leckie, C. and Ramamohanarao, K. and Palaniswami, M. , title =. IEEE Trans. Fuzzy Syst. , volume =. 2013 , pages =

    work page 2013

  25. [25]

    Newman, M. E. J. , title =

    work page

  26. [26]

    2012 , pages =

    Ubiquitousness of link-density and link-pattern communities in real-world networks , journal =. 2012 , pages =

    work page 2012

  27. [27]

    and Fortunato, S

    Lancichinetti, A. and Fortunato, S. and Radicchi, F. , title =. Phys. Rev. E. , volume =. 2008 , pages =

    work page 2008

  28. [28]

    and Pellegrini, M

    Liu, W. and Pellegrini, M. and Wang, X. , title =. Sci. Rep. , volume =. 2014 , pages =

    work page 2014

  29. [29]

    and Diaz-Guilera, A

    Danon, L. and Diaz-Guilera, A. and Duch, J. and Arenas, A. , title =. J. Stat. Mech.-Theory Exp. , volume =. 2005 , pages =

    work page 2005

  30. [30]

    , title =

    Gregory, S. , title =. J. Stat. Mech.-Theory Exp. , volume =. 2011 , pages =

    work page 2011

  31. [31]

    and Fortunato, S

    Lancichinetti, A. and Fortunato, S. , title =. Phys. Rev. E. , volume =. 2009 , pages =

    work page 2009

  32. [32]

    and Rifqi, M

    Hullermeier, E. and Rifqi, M. , title =. in Proc. IFSA/EUSFLAT Conf. , year =

    work page

  33. [33]

    Remote sensing of environment , volume=

    Lidar remote sensing of the canopy structure and biophysical properties of Douglas-fir western hemlock forests , author=. Remote sensing of environment , volume=. 1999 , publisher=

    work page 1999

  34. [34]

    ISPRS Journal of photogrammetry and remote sensing , volume=

    Airborne laser scanning—an introduction and overview , author=. ISPRS Journal of photogrammetry and remote sensing , volume=. 1999 , publisher=

    work page 1999

  35. [35]

    ISPRS Journal of Photogrammetry and remote sensing , volume=

    Predicting individual tree attributes from airborne laser point clouds based on the random forests technique , author=. ISPRS Journal of Photogrammetry and remote sensing , volume=. 2011 , publisher=

    work page 2011

  36. [36]

    Remote Sensing of Environment , volume=

    Individual tree species classification using low-density airborne multispectral LiDAR data via attribute-aware cross-branch transformer , author=. Remote Sensing of Environment , volume=. 2024 , publisher=

    work page 2024

  37. [37]

    European Journal of Remote Sensing , volume=

    Quantifying changes in forest structural complexity using bi-temporal airborne laser scanning , author=. European Journal of Remote Sensing , volume=. 2024 , publisher=

    work page 2024

  38. [38]

    Science of Remote Sensing , volume=

    Direct and automatic measurements of stem curve and volume using a high-resolution airborne laser scanning system , author=. Science of Remote Sensing , volume=. 2022 , publisher=

    work page 2022

  39. [39]

    4th ISPRS Geospatial Week 2019 , pages=

    Semantic labeling of ALS point clouds for tree species mapping using the deep neural network PointNet++ , author=. 4th ISPRS Geospatial Week 2019 , pages=. 2019 , organization=

    work page 2019

  40. [40]

    ISPRS annals of the photogrammetry, remote sensing and spatial information sciences , volume=

    Classification of tree species and standing dead trees by fusing UAV-based lidar data and multispectral imagery in the 3D deep neural network PointNet++ , author=. ISPRS annals of the photogrammetry, remote sensing and spatial information sciences , volume=. 2020 , publisher=

    work page 2020

  41. [41]

    ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences , volume=

    Detection of single tree stems in forested areas from high density ALS point clouds using 3D shape descriptors , author=. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences , volume=. 2017 , publisher=

    work page 2017

  42. [42]

    Remote Sensing of Environment , volume=

    Deep point cloud regression for above-ground forest biomass estimation from airborne LiDAR , author=. Remote Sensing of Environment , volume=. 2024 , publisher=

    work page 2024

  43. [43]

    Photogrammetric Engineering & Remote Sensing , volume=

    A new method for segmenting individual trees from the lidar point cloud , author=. Photogrammetric Engineering & Remote Sensing , volume=. 2012 , publisher=

    work page 2012

  44. [44]

    Proceedings of the 30th international conference on advances in geographic information systems , pages=

    Deep learning based 3D point cloud regression for estimating forest biomass , author=. Proceedings of the 30th international conference on advances in geographic information systems , pages=

    work page

  45. [45]

    IEEE Transactions on visualization and computer graphics , volume=

    Computing and rendering point set surfaces , author=. IEEE Transactions on visualization and computer graphics , volume=. 2003 , publisher=

    work page 2003

  46. [46]

    ACM Transactions on Graphics (ToG) , volume=

    Parameterization-free projection for geometry reconstruction , author=. ACM Transactions on Graphics (ToG) , volume=. 2007 , publisher=

    work page 2007

  47. [47]

    ACM transactions on graphics (TOG) , volume=

    Consolidation of unorganized point clouds for surface reconstruction , author=. ACM transactions on graphics (TOG) , volume=. 2009 , publisher=

    work page 2009

  48. [48]

    ACM transactions on graphics (TOG) , volume=

    Edge-aware point set resampling , author=. ACM transactions on graphics (TOG) , volume=. 2013 , publisher=

    work page 2013

  49. [49]

    Science of The Total Environment , volume=

    Species discrimination and individual tree detection for predicting main dendrometric characteristics in mixed temperate forests by use of airborne laser scanning and ultra-high-resolution imagery , author=. Science of The Total Environment , volume=. 2020 , publisher=

    work page 2020

  50. [50]

    Remote Sensing , volume=

    Comparative analysis of multi-platform, multi-resolution, multi-temporal LiDAR data for forest inventory , author=. Remote Sensing , volume=. 2022 , publisher=

    work page 2022

  51. [51]

    Proceedings of the IEEE conference on computer vision and pattern recognition , pages=

    Pu-net: Point cloud upsampling network , author=. Proceedings of the IEEE conference on computer vision and pattern recognition , pages=

    work page

  52. [52]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Grad-pu: Arbitrary-scale point cloud upsampling via gradient descent with learned distance functions , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    work page

  53. [53]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Point cloud upsampling via disentangled refinement , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    work page

  54. [54]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Patch-based progressive 3d point set upsampling , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    work page

  55. [55]

    Proceedings of the IEEE/CVF international conference on computer vision , pages=

    Pu-gan: a point cloud upsampling adversarial network , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=

    work page

  56. [56]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    RepKPU: Point cloud upsampling with kernel point representation and deformation , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    work page

  57. [57]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    A conditional denoising diffusion probabilistic model for point cloud upsampling , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    work page

  58. [58]

    arXiv preprint arXiv:2501.15286 , year=

    Efficient Point Clouds Upsampling via Flow Matching , author=. arXiv preprint arXiv:2501.15286 , year=

    work page arXiv

  59. [59]

    Canadian Journal of Remote Sensing , volume=

    Remote sensing technologies for enhancing forest inventories: A review , author=. Canadian Journal of Remote Sensing , volume=. 2016 , publisher=

    work page 2016

  60. [60]

    Remote Sensing , volume=

    Tree Completion Net: A Novel Vegetation Point Clouds Completion Model Based on Deep Learning , author=. Remote Sensing , volume=. 2024 , publisher=

    work page 2024

  61. [61]

    Remote Sensing , volume=

    Restoration of individual tree missing point cloud based on local features of point cloud , author=. Remote Sensing , volume=. 2022 , publisher=

    work page 2022

  62. [62]

    Electronics , volume=

    A single-tree point cloud completion approach of feature fusion for agricultural robots , author=. Electronics , volume=. 2023 , publisher=

    work page 2023

  63. [63]

    Science , volume=

    The biodiversity and ecosystem service contributions and trade-offs of forest restoration approaches , author=. Science , volume=. 2022 , publisher=

    work page 2022

  64. [64]

    science , volume=

    A large and persistent carbon sink in the world’s forests , author=. science , volume=. 2011 , publisher=

    work page 2011

  65. [65]

    Science , volume=

    A climate risk analysis of Earth’s forests in the 21st century , author=. Science , volume=. 2022 , publisher=

    work page 2022

  66. [66]

    Science , volume=

    Climate-driven risks to the climate mitigation potential of forests , author=. Science , volume=. 2020 , publisher=

    work page 2020

  67. [67]

    ISPRS journal of photogrammetry and remote sensing , volume=

    Scanning geometry: Influencing factor on the quality of terrestrial laser scanning points , author=. ISPRS journal of photogrammetry and remote sensing , volume=. 2011 , publisher=

    work page 2011

  68. [68]

    Science of Remote Sensing , volume=

    Large-scale inventory in natural forests with mobile LiDAR point clouds , author=. Science of Remote Sensing , volume=. 2024 , publisher=

    work page 2024

  69. [69]

    SilviLaser 2021 Benchmark Dataset - Terrestrial Challenge , DOI=

    Hollaus, Markus and Chen, Yi-Chen , year=. SilviLaser 2021 Benchmark Dataset - Terrestrial Challenge , DOI=

    work page 2021

  70. [70]

    IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium , pages=

    Integrating Airborne and Terrestrial Laser Scanning for Complete 3D Model Generation in Dense Forest , author=. IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium , pages=. 2023 , organization=

    work page 2023

  71. [71]

    arXiv preprint arXiv:2410.17001 , year=

    Joint Point Cloud Upsampling and Cleaning with Octree-based CNNs , author=. arXiv preprint arXiv:2410.17001 , year=

    work page arXiv

  72. [72]

    Global change biology , volume=

    Allometric equations for integrating remote sensing imagery into forest monitoring programmes , author=. Global change biology , volume=. 2017 , publisher=

    work page 2017

  73. [73]

    Computer graphics and image processing , volume=

    Geometric modeling using octree encoding , author=. Computer graphics and image processing , volume=. 1982 , publisher=

    work page 1982

  74. [74]

    2018 , publisher =

    Parkan, Matthew , title =. 2018 , publisher =. doi:10.5281/zenodo.1252955 , url =

    work page doi:10.5281/zenodo.1252955 2018

  75. [75]

    Earth System Science Data , volume=

    Individual tree point clouds and tree measurements from multi-platform laser scanning in German forests , author=. Earth System Science Data , volume=. 2022 , publisher=

    work page 2022

  76. [76]

    Laser Radar Technology and Applications XX; and Atmospheric Propagation XII , volume=

    Medium altitude airborne Geiger-mode mapping LIDAR system , author=. Laser Radar Technology and Applications XX; and Atmospheric Propagation XII , volume=. 2015 , organization=

    work page 2015

  77. [77]

    Laser radar technology and applications xxi , volume=

    Linear LIDAR versus Geiger-mode LIDAR: impact on data properties and data quality , author=. Laser radar technology and applications xxi , volume=. 2016 , organization=

    work page 2016

  78. [78]

    Remote Sensing , volume=

    Evaluation of single photon and Geiger mode LiDAR for the 3D elevation program , author=. Remote Sensing , volume=. 2016 , publisher=

    work page 2016

  79. [79]

    Remote Sensing of Environment , volume=

    A three-stage framework for stand-level automated stem volume estimation in temperate forests using Mobile laser scanning , author=. Remote Sensing of Environment , volume=. 2026 , publisher=

    work page 2026

  80. [80]

    ISPRS journal of photogrammetry and remote sensing , volume=

    Terrestrial laser scanning in forest inventories , author=. ISPRS journal of photogrammetry and remote sensing , volume=. 2016 , publisher=

    work page 2016

Showing first 80 references.