pith. sign in

arxiv: 2604.03328 · v1 · submitted 2026-04-02 · 💻 cs.CV · cs.RO

Review and Evaluation of Point-Cloud based Leaf Surface Reconstruction Methods for Agricultural Applications

Arif Ahmed , Parikshit Maini This is my paper

Pith reviewed 2026-05-13 21:11 UTC · model grok-4.3

classification 💻 cs.CV cs.RO
keywords leaf surface reconstructionpoint cloudsurface reconstructionagricultural roboticsphenotypingcomparative evaluation3D reconstructionresource constraints
0
0 comments X

The pith

Nine point-cloud surface reconstruction methods for leaves each excel in different agricultural scenarios based on accuracy, noise robustness, and computation needs.

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

The paper conducts a comparative evaluation of nine surface reconstruction methods on three diverse leaf point-cloud datasets. It measures performance in terms of surface area accuracy, smoothness, handling of noise and missing data, and processing time. The study finds that no method dominates universally; instead, choices depend on the specific application and the robotic platform's resource limits. This matters because accurate leaf models support plant phenotyping and decision-making in farming, where hardware is often constrained. By testing indoor high-res to field noisy scans, the results offer selection criteria for practitioners.

Core claim

The central claim is that a systematic comparison of nine representative methods—spanning parametric, triangulation, implicit, and learning-based categories—on datasets covering clean indoor and noisy field data reveals distinct advantages for each technique. Performance metrics trade off surface fidelity against computational demands, enabling informed selection for resource-limited agricultural robots rather than assuming one best method.

What carries the argument

The multi-metric evaluation across accuracy, smoothness, robustness, and cost on three public datasets serves as the mechanism to expose trade-offs between reconstruction approaches.

If this is right

  • Robotic systems in fields can prioritize faster, more robust methods when data is noisy.
  • Indoor phenotyping setups may favor higher-accuracy but slower techniques.
  • Method choice directly impacts the feasibility of real-time leaf analysis on limited hardware.
  • General guidance emerges for balancing reconstruction quality with deployment costs in agriculture.

Where Pith is reading between the lines

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

  • Future work could test these methods on additional crop types to check if the observed trade-offs persist.
  • Integrating the best-performing method per scenario into robotic pipelines might improve overall phenotyping throughput.
  • Extending the evaluation to dynamic scenes with moving leaves could reveal new robustness requirements.

Load-bearing premise

The nine chosen methods adequately represent the range of current techniques and that the performance patterns seen on the tested datasets will hold for other leaves, sensors, and environments.

What would settle it

A new evaluation on a fourth dataset with different leaf morphology or sensor noise profile that shows the same methods ranking differently or uniform performance across methods would undermine the claim of distinct advantages.

Figures

Figures reproduced from arXiv: 2604.03328 by Arif Ahmed, Parikshit Maini.

Figure 1
Figure 1. Figure 1: Leaf surface Reconstruction results comparison across methods that we described in Sec. 4 4.8. Ball Pivoting Algorithm (BPA) This surface reconstruction method creates triangular meshes from point cloud by simulating a ball of radius 𝑟 rolling over points Bernardini, Mittleman, Rushmeier, Silva and Taubin (2002). The algorithm works by pivoting a virtual ball of radius 𝑟 across the point cloud, where the r… view at source ↗
Figure 2
Figure 2. Figure 2: Mean leaf surface area (%) for each plant computed by different methods in Sec. 4 on   dataset using Poisson as benchmark [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Reconstruction of Trifoliate Strawberry Leaf Clusters Figure 4b that SOM exhibits the smallest IQR, indicating that variability in leaf point clouds has a limited effect on its peak RAM consumption. In [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Comparison of Leaf Surface Reconstruction Methods on the   Dataset (a) Mean CPU Time per Plant (b) Variability of Peak RAM Usage (IQR) creating triangles that fill in regions that should remain open or concave. We observe that BPA, shown in the top right of [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
read the original abstract

Accurate reconstruction of leaf surfaces from 3D point cloud is essential for agricultural applications such as phenotyping. However, real-world plant data (i.e., irregular 3D point cloud) are often complex to reconstruct plant parts accurately. A wide range of surface reconstruction methods has been proposed, including parametric, triangulation-based, implicit, and learning based approaches, yet their relative performance for leaf surface reconstruction remains insufficiently understood. In this work, we present a comparative study of nine representative surface reconstruction methods for leaf surfaces. We evaluate these methods on three publicly available datasets: LAST-STRAW, Pheno4D, and Crops3D - spanning diverse species, sensors, and sensing environments, ranging from clean high-resolution indoor scans to noisy low-resolution field settings. The analysis highlights the trade-offs between surface area estimation accuracy, smoothness, robustness to noise and missing data, and computational cost across different methods. These factors affect the cost and constraints of robotic hardware used in agricultural applications. Our results show that each method exhibits distinct advantages depending on application and resource constraints. The findings provide practical guidance for selecting surface reconstruction techniques for resource constrained robotic platforms.

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

2 major / 3 minor

Summary. The manuscript presents a comparative evaluation of nine point-cloud surface reconstruction methods (parametric, triangulation-based, implicit, and learning-based) for leaf surfaces in agricultural phenotyping and robotics. The methods are tested on three public datasets—LAST-STRAW, Pheno4D, and Crops3D—spanning indoor high-resolution scans to noisy field data across species and sensors. Evaluation uses metrics for surface area accuracy, smoothness, noise/missing-data robustness, and computational cost. The central claim is that the methods exhibit distinct trade-offs, enabling practical guidance for method selection under resource constraints on robotic platforms.

Significance. If the observed trade-offs are robust, the work provides a valuable empirical benchmark that addresses the lack of targeted comparisons for leaf surfaces (as opposed to generic 3D objects). Credit is due for the use of multiple public datasets, relevant application-specific metrics, and explicit attention to computational cost, which directly supports deployment decisions on resource-constrained hardware. This can inform efficient 3D perception pipelines in agricultural robotics and phenotyping.

major comments (2)
  1. [Datasets and Experimental Setup] Datasets section: The practical guidance claim rests on the representativeness of the three datasets. While they cover some species/sensor variation, the manuscript provides no quantitative analysis of coverage (e.g., leaf morphology diversity, occlusion statistics, or sensor artifact distributions) or sensitivity tests showing that relative rankings remain stable under additional field conditions. This is load-bearing for the central claim.
  2. [Results and Discussion] Results section: Performance differences are presented without statistical significance tests (e.g., paired t-tests or ANOVA with multiple-comparison correction) or confidence intervals on metrics such as surface area error. This makes it difficult to confirm that reported 'distinct advantages' are reliable rather than within noise.
minor comments (3)
  1. [Abstract] Abstract: The phrasing 'real-world plant data (i.e., irregular 3D point cloud) are often complex to reconstruct plant parts accurately' is awkward; rephrase for clarity.
  2. [Methods] Methods: Provide version numbers, implementation sources (e.g., GitHub links or library versions), and hyperparameter settings for each of the nine methods to ensure reproducibility.
  3. [Figures] Figures: Captions for reconstruction visualizations should include scale information and specify the exact viewpoint or rendering parameters used.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and the recommendation for minor revision. We address each major comment point by point below.

read point-by-point responses

  1. Referee: Datasets section: The practical guidance claim rests on the representativeness of the three datasets. While they cover some species/sensor variation, the manuscript provides no quantitative analysis of coverage (e.g., leaf morphology diversity, occlusion statistics, or sensor artifact distributions) or sensitivity tests showing that relative rankings remain stable under additional field conditions. This is load-bearing for the central claim.

    Authors: We agree that additional quantitative characterization of dataset coverage would strengthen support for the practical guidance claims. In the revised manuscript we will add: (i) summary statistics on leaf morphology diversity (area, aspect ratio, and curvature distributions per species), (ii) occlusion proxies derived from point-density variation, and (iii) sensor-artifact profiles (noise level and missing-data fraction) across the three datasets. Full sensitivity tests under entirely new field conditions would require fresh data acquisition and are outside the scope of this minor revision; we will instead expand the discussion to state the coverage limitations and the conditions under which the reported trade-offs are expected to generalize. revision: partial

  2. Referee: Results section: Performance differences are presented without statistical significance tests (e.g., paired t-tests or ANOVA with multiple-comparison correction) or confidence intervals on metrics such as surface area error. This makes it difficult to confirm that reported 'distinct advantages' are reliable rather than within noise.

    Authors: We acknowledge the value of statistical validation. In the revised version we will add paired t-tests (or Wilcoxon signed-rank tests where normality assumptions are violated) with appropriate multiple-comparison correction (Bonferroni or FDR) for the primary metrics, together with 95 % confidence intervals or standard errors on surface-area error and other key quantities. These additions will be presented in updated tables and figures. revision: yes

Circularity Check

0 steps flagged

No circularity: purely empirical comparison of existing methods

full rationale

The paper performs a comparative evaluation of nine pre-existing surface reconstruction algorithms on three independent public datasets (LAST-STRAW, Pheno4D, Crops3D). No derivations, fitted parameters, predictions, or self-referential equations appear; all reported advantages and trade-offs are direct observations from external data. Self-citations, if present, are not load-bearing for the central claims, which rest on reproducible metrics rather than internal definitions or ansatzes.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

This is an empirical comparative review with no mathematical derivations or new postulates. It relies on standard computer vision assumptions about point cloud data and surface reconstruction applicability to leaves.

axioms (1)
  • domain assumption The three chosen datasets adequately represent the range of noise, resolution, and environmental conditions encountered in agricultural leaf scanning
    Invoked to support claims about robustness across indoor and field settings.

pith-pipeline@v0.9.0 · 5501 in / 1278 out tokens · 64779 ms · 2026-05-13T21:11:20.425086+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

    work page

  2. [2]

    , author Agarwal, R

    author Ahmed, A. , author Agarwal, R. , author Srikar, G. , author Rose, N. , author Maini, P. , year 2025 . title Saral-bot: Autonomous robot for strawberry plant care . journal arXiv preprint arXiv:2506.06798

    work page arXiv 2025

  3. [3]

    , author Ozasa, Y

    author Ando, R. , author Ozasa, Y. , author Guo, W. , year 2021 . title Robust surface reconstruction of plant leaves from 3d point clouds . journal Plant Phenomics

    work page 2021

  4. [4]

    , author Mittleman, J

    author Bernardini, F. , author Mittleman, J. , author Rushmeier, H. , author Silva, C. , author Taubin, G. , year 2002 . title The ball-pivoting algorithm for surface reconstruction . journal IEEE transactions on visualization and computer graphics volume 5 , pages 349--359

    work page 2002

  5. [5]

    , author Krishnamurthy, A

    author Bingol, O.R. , author Krishnamurthy, A. , year 2019 . title NURBS-Python : An open-source object-oriented NURBS modeling framework in Python . journal SoftwareX volume 9 , pages 85--94 . :https://doi.org/10.1016/j.softx.2018.12.005

    work page doi:10.1016/j.softx.2018.12.005 2019

  6. [6]

    , author van Henten, E.J

    author Burusa, A.K. , author van Henten, E.J. , author Kootstra, G. , year 2024 a. title Attention-driven next-best-view planning for efficient reconstruction of plants and targeted plant parts . journal Biosystems Engineering volume 246 , pages 248--262

    work page 2024

  7. [7]

    , author van Henten, E.J

    author Burusa, A.K. , author van Henten, E.J. , author Kootstra, G. , year 2024 b. title Gradient-based local next-best-view planning for improved perception of targeted plant nodes , in: booktitle 2024 IEEE International Conference on Robotics and Automation (ICRA) , organization IEEE . pp. pages 15854--15860

    work page 2024

  8. [8]

    , author van Henten, E.J

    author Ci, J. , author van Henten, E.J. , author Wang, X. , author Burusa, A.K. , author Kootstra, G. , year 2025 . title Ssl-nbv: A self-supervised-learning-based next-best-view algorithm for efficient 3d plant reconstruction by a robot . journal Computers and Electronics in Agriculture volume 233 , pages 110121

    work page 2025

  9. [9]

    , author Devlin, S.J

    author Cleveland, W.S. , author Devlin, S.J. , year 1988 . title Locally weighted regression: an approach to regression analysis by local fitting . journal Journal of the American statistical association volume 83 , pages 596--610

    work page 1988

  10. [10]

    , author Yilmaz, A

    author Dang, R. , author Yilmaz, A. , author Cielniak, G. , year 2025 . title Plant2sim3d: From real plants to simulated worlds via video-driven 3d mesh generation

    work page 2025

  11. [11]

    , author Cheong, O

    author De Berg, M. , author Cheong, O. , author Van Kreveld, M. , author Overmars, M. , year 2008 . title Computational geometry: algorithms and applications . publisher Springer

    work page 2008

  12. [12]

    , author Gu, R

    author Dimitrov, A. , author Gu, R. , author Golparvar-Fard, M. , year 2016 . title Non-uniform b-spline surface fitting from unordered 3d point clouds for as-built modeling . journal Computer-Aided Civil and Infrastructure Engineering volume 31 , pages 483--498

    work page 2016

  13. [13]

    , author Kempthorne, D.M

    author Dorr, G.J. , author Kempthorne, D.M. , author Mayo, L.C. , author Forster, W.A. , author Zabkiewicz, J.A. , author McCue, S.W. , author Belward, J.A. , author Turner, I.W. , author Hanan, J. , year 2014 . title Towards a model of spray--canopy interactions: interception, shatter, bounce and retention of droplets on horizontal leaves . journal Ecolo...

    work page 2014

  14. [14]

    , year 1988

    author Field, D.A. , year 1988 . title Laplacian smoothing and delaunay triangulations . journal Communications in applied numerical methods volume 4 , pages 709--712

    work page 1988

  15. [15]

    , author Cohen-Or, D

    author Fleishman, S. , author Cohen-Or, D. , author Silva, C.T. , year 2005 . title Robust moving least-squares fitting with sharp features . journal ACM transactions on graphics (TOG) volume 24 , pages 544--552

    work page 2005

  16. [16]

    , author Hofer, M

    author Fl \"o ry, S. , author Hofer, M. , year 2008 . title Constrained curve fitting on manifolds . journal Computer-Aided Design volume 40 , pages 25--34

    work page 2008

  17. [17]

    , author He, Y

    author Hui, Z. , author He, Y. , author Jin, S. , author Chen, W. , author He, H. , author Ziggah, Y.Y. , year 2026 . title Leaflods: A self-adaptive 3-d leaf modeling with enhancing level of details expression . journal Computers and Electronics in Agriculture volume 243 , pages 111377

    work page 2026

  18. [18]

    , author Pan, S

    author Isaac Jose, A. , author Pan, S. , author Zaenker, T. , author Menon, R. , author Houben, S. , author Bennewitz, M. , year 2025 . title GO-VMP : G lobal optimization for view motion planning in fruit mapping , in: booktitle IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    work page 2025

  19. [19]

    , author Heiwolt, K

    author James, K.M.F. , author Heiwolt, K. , author Sargent, D.J. , author Cielniak, G. , year 2024 . title Lincoln's Annotated Spatio-Temporal Strawberry Dataset (LAST-Straw) . arXiv:2403.00566 http://arxiv.org/abs/2403.00566

    work page arXiv 2024

  20. [20]

    , author Bolitho, M

    author Kazhdan, M. , author Bolitho, M. , author Hoppe, H. , year 2006 . title Poisson surface reconstruction , in: booktitle Proceedings of the fourth Eurographics symposium on Geometry processing

    work page 2006

  21. [21]

    , author Turner, I.W

    author Kempthorne, D.M. , author Turner, I.W. , author Belward, J.A. , author McCue, S.W. , author Barry, M. , author Young, J. , author Dorr, G.J. , author Hanan, J. , author Zabkiewicz, J.A. , year 2015 . title Surface reconstruction of wheat leaf morphology from three-dimensional scanned data . journal Functional Plant Biology volume 42 , pages 444--451

    work page 2015

  22. [22]

    , year 1997

    author Kohonen, T. , year 1997 . title Exploration of very large databases by self-organizing maps , in: booktitle Proceedings of international conference on neural networks (icnn'97) , organization IEEE . pp. pages PL1--PL6

    work page 1997

  23. [23]

    , author Liang, Z

    author Li, Y. , author Liang, Z. , author Liu, B. , author Yin, L. , author Wan, F. , author Qian, W. , author Qiao, X. , year 2025 . title Applications of 3d reconstruction techniques in crop canopy phenotyping: A review. journal Agronomy volume 15

    work page 2025

  24. [24]

    , author Haron, H

    author Lim, S.P. , author Haron, H. , year 2014 . title Surface reconstruction techniques: a review . journal Artificial Intelligence Review volume 42 , pages 59--78

    work page 2014

  25. [25]

    , author Belward, J

    author Loch, B.I. , author Belward, J. , author Hanan, J. , year 2005 . title Application of surface fitting techniques for the representation of leaf surfaces

    work page 2005

  26. [26]

    , year 2021

    author Masuda, T. , year 2021 . title Leaf area estimation by semantic segmentation of point cloud of tomato plants , in: booktitle Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. pages 1381--1389

    work page 2021

  27. [27]

    , author McCue, S.W

    author Mayo, L.C. , author McCue, S.W. , author Moroney, T.J. , author Forster, W.A. , author Kempthorne, D.M. , author Belward, J.A. , author Turner, I.W. , year 2015 . title Simulating droplet motion on virtual leaf surfaces . journal Royal Society open science volume 2

    work page 2015

  28. [28]

    , author Zaenker, T

    author Menon, R. , author Zaenker, T. , author Dengler, N. , author Bennewitz, M. , year 2023 . title Nbv-sc: Next best view planning based on shape completion for fruit mapping and reconstruction , in: booktitle 2023 IEEE/RSJ international conference on intelligent robots and systems (IROS) , organization IEEE . pp. pages 4197--4203

    work page 2023

  29. [29]

    , author Tiller, W

    author Piegl, L. , author Tiller, W. , year 2012 . title The NURBS book . publisher Springer Science & Business Media

    work page 2012

  30. [30]

    , author Tan, P

    author Quan, L. , author Tan, P. , author Zeng, G. , author Yuan, L. , author Wang, J. , author Kang, S.B. , year 2006 . title Image-based plant modeling , in: booktitle ACM Siggraph 2006 Papers , pp. pages 599--604

    work page 2006

  31. [31]

    , author Isler, V

    author Roy, P. , author Isler, V. , year 2017 . title Active view planning for counting apples in orchards , in: booktitle 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , organization IEEE . pp. pages 6027--6032

    work page 2017

  32. [32]

    , author Silva, L.A

    author Sampaio, G.S. , author Silva, L.A. , author Marengoni, M. , year 2021 . title 3d reconstruction of non-rigid plants and sensor data fusion for agriculture phenotyping . journal Sensors volume 21 , pages 4115

    work page 2021

  33. [33]

    , author Koenigkan, L.V

    author Santos, T.T. , author Koenigkan, L.V. , author Barbedo, J.G.A. , author Rodrigues, G.C. , year 2014 . title 3d plant modeling: localization, mapping and segmentation for plant phenotyping using a single hand-held camera , in: booktitle European Conference on Computer Vision , organization Springer . pp. pages 247--263

    work page 2014

  34. [34]

    , author Magistri, F

    author Schunck, D. , author Magistri, F. , author Rosu, R.A. , author Corneli en, A. , author Chebrolu, N. , author Paulus, S. , author L \'e on, J. , author Behnke, S. , author Stachniss, C. , author Kuhlmann, H. , et al., year 2021 . title Pheno4d: A spatio-temporal dataset of maize and tomato plant point clouds for phenotyping and advanced plant analys...

    work page 2021

  35. [35]

    , author Jost, B

    author Stausberg, L. , author Jost, B. , author Klingbeil, L. , author Kuhlmann, H. , year 2024 . title A 3d surface reconstruction pipeline for plant phenotyping . journal Remote Sensing volume 16 , pages 4720

    work page 2024

  36. [36]

    , author Lu, L

    author Wang, L. , author Lu, L. , author Jiang, N. , year 2011 . title A study of leaf modeling technology based on morphological features . journal Mathematical and computer modelling volume 54 , pages 1107--1114

    work page 2011

  37. [37]

    , author Li, L

    author Wang, X. , author Li, L. , author Chai, W. , year 2013 . title Geometric modeling of broad-leaf plants leaf based on b-spline . journal Mathematical and computer Modelling volume 58 , pages 564--572

    work page 2013

  38. [38]

    , author Liu, S

    author Wei, K. , author Liu, S. , author Chen, Q. , author Huang, S. , author Zhong, M. , author Zhang, J. , author Sun, H. , author Wu, K. , author Fan, S. , author Ye, Z. , et al., year 2024 . title Fast multi-view 3d reconstruction of seedlings based on automatic viewpoint planning . journal Computers and Electronics in Agriculture volume 218 , pages 108708

    work page 2024

  39. [39]

    , author Wen, W

    author Wu, S. , author Wen, W. , author Gou, W. , author Lu, X. , author Zhang, W. , author Zheng, C. , author Xiang, Z. , author Chen, L. , author Guo, X. , year 2022 . title A miniaturized phenotyping platform for individual plants using multi-view stereo 3d reconstruction . journal Frontiers in plant science volume 13 , pages 897746

    work page 2022

  40. [40]

    , author Hu, Y

    author Wu, W. , author Hu, Y. , author Lu, Y. , year 2021 . title Parametric surface modelling for tea leaf point cloud based on non-uniform rational basis spline technique . journal Sensors volume 21 , pages 1304

    work page 2021

  41. [41]

    , author Morellas, V

    author Zermas, D. , author Morellas, V. , author Mulla, D. , author Papanikolopoulos, N. , year 2017 . title Estimating the leaf area index of crops through the evaluation of 3d models , in: booktitle 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , organization IEEE . pp. pages 6155--6162

    work page 2017

  42. [42]

    , author Morellas, V

    author Zermas, D. , author Morellas, V. , author Mulla, D. , author Papanikolopoulos, N. , year 2020 . title 3d model processing for high throughput phenotype extraction--the case of corn . journal Computers and Electronics in Agriculture volume 172 , pages 105047

    work page 2020

  43. [43]

    , author Osher, S

    author Zhao, H.K. , author Osher, S. , author Fedkiw, R. , year 2001 . title Fast surface reconstruction using the level set method , in: booktitle 1st IEEE Workshop on Variational and Level Set Methods , pp. pages 194--202

    work page 2001

  44. [44]

    , author Thapa, S

    author Zhu, F. , author Thapa, S. , author Gao, T. , author Ge, Y. , author Walia, H. , author Yu, H. , year 2018 . title 3d reconstruction of plant leaves for high-throughput phenotyping , in: booktitle 2018 IEEE International Conference on Big Data (Big Data) , organization IEEE . pp. pages 4285--4293

    work page 2018

  45. [45]

    , author Zhai, R

    author Zhu, J. , author Zhai, R. , author Ren, H. , author Xie, K. , author Du, A. , author He, X. , author Cui, C. , author Wang, Y. , author Ye, J. , author Wang, J. , et al., year 2024 . title Crops3d: a diverse 3d crop dataset for realistic perception and segmentation toward agricultural applications . journal Scientific Data volume 11 , pages 1438

    work page 2024