Pith. sign in

REVIEW 4 major objections 6 minor 59 references

A Low-Cost Machine Learning Approach for Timber Diameter Estimation

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper establishes that a fine-tuned YOLOv5 detector on ordinary RGB images can localize timber logs and assign diameter ranges via bounding-box width, with a reported test mAP@0.5 of 0.64.

desk verdict A clearly written application note whose central diameter-estimation claim is circular: labels and predictions both reduce to bounding-box width, and Section 4 reports no bin accuracy or physical comparison. read the letter →

arxiv 2507.17219 v1 pith:Y5QYAUXR submitted 2025-07-23 cs.CV cs.LG

Fatemeh Hasanzadeh Fard , Sanaz Hasanzadeh Fard , Mehdi Jonoobi This is my paper
classification cs.CVcs.LG
keywords timberdetectiondiameterestimationYOLOv5computervisionwoodprocessingmachinelearningSeg1.0object
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that a standard YOLOv5 model, fine-tuned on the public TimberSeg 1.0 dataset of 440 images with 10,735 annotated logs, can detect timber logs in ordinary RGB photographs taken in industrial sheds. The point of the work is to show that a low-cost, CPU-runnable pipeline can automate log detection and produce a rough diameter estimate, replacing slow and inconsistent manual inspection in small- and medium-sized wood-processing operations. On a held-out test set, the model reports a mean average precision (mAP@0.5) of 0.64, precision of 0.656, and per-image CPU inference under 300 ms. Diameter is estimated in post-processing: the width of each detected bounding box is binned into Thin, Medium, or Thick categories. If the claim holds, this offers an affordable first-pass sorting and inventory tool that needs no laser scanners, 3D cameras, or controlled lighting.

What carries the argument

The carrying mechanism is YOLOv5, a single-shot convolutional object detector that predicts bounding boxes and class labels in one pass, combined with transfer learning from pretrained weights. The identity that does the estimation work is the bounding-box width of each detection, converted into three diameter bins (Thin <30 px, Medium 30–60 px, Thick >60 px) in a post-processing step. Because the training labels themselves are defined by the same width thresholds, the quality of the diameter assignment is directly tied to the quality of the detector's localization. Training uses images resized to 416×416 with batch sizes of 8–16, and runs 50–100 epochs with early stopping.

What would settle it

Take a set of logs with hand-measured diameters, photograph each from several distances and angles in a sawmill yard, run the fine-tuned model, and compute bin accuracy separately by distance and angle. If bin accuracy does not stay consistently high across those conditions, the bounding-box-width proxy for diameter fails.

Watch

Extended reading notes

Core claim

The central discovery is that a fine-tuned YOLOv5 detector, trained for only 50 epochs in a CPU-only environment on a small single-class dataset, can reliably localize logs in unseen images and sort them into rough diameter categories defined by bounding-box pixel width. On the test set the model achieves precision of 0.656, recall of 0.577, mAP@0.5 of 0.640, and mAP@0.5:0.95 of 0.356, with per-image CPU inference under 300 ms. The diameter estimate is not a separate measurement stage: it is a post-processing step that reads the width of each detected box and assigns a bin (Thin <30 px, Medium 30–60 px, Thick >60 px). The paper frames this as an application-oriented contribution, showing that existing tools can deliver practical value in real-world timber yards without specialized hardware.

Load-bearing premise

The load-bearing premise is that the width of the box drawn around each detected log reflects the log's real diameter no matter the camera distance, viewing angle, or partial hiding; the paper does not independently measure true diameters to test this.

Editorial extensions

If this is right

  • A CPU-only fine-tuning run on a small, single-class dataset can converge and detect logs in unseen images with mAP@0.5 of 0.64.
  • Diameter can be assigned in post-processing by binning bounding-box width into Thin, Medium, and Thick, giving a first-pass sort without extra sensors.
  • Per-image CPU inference under 300 ms makes the pipeline usable for near-real-time inventory scanning or conveyor-side sorting.
  • The same detector could be extended with additional classes for wood species and quality, and with depth input for better measurements, as the paper notes in its future-work discussion.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper's own claims, a natural next test is to compare the binned estimates against hand-measured log diameters; the paper does not report such a comparison, so the reported accuracy measures consistency with pixel-size bins rather than with physical diameter.
  • A per-site calibration could make bounding-box width more reliable: placing an object of known size in the frame would convert pixel widths to approximate physical diameters without adding depth sensors.
  • If combined with log length and stack geometry from the same images, the width bins could support approximate volume estimates, a step the current work does not attempt.
  • The single-class setup treats all logs alike; adding species and quality classes would turn the detector from a sorting aid into a more complete inventory tool.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This manuscript presents a low-cost pipeline for detecting timber logs in RGB images and assigning them to coarse diameter categories. The authors fine-tune a YOLOv5 detector on the public TimberSeg 1.0 dataset using a CPU-only Google Colab environment, then assign diameter ranges from bounding-box width as a post-processing step. On a held-out test set of 21 images containing 208 labeled instances, they report precision 0.656, recall 0.577, mAP@0.5 0.640, and mAP@0.5:0.95 0.356. The abstract and conclusion claim that the model estimates timber diameter from bounding-box dimensions for practical sorting and inventory use.

Significance. If the diameter-estimation claim were substantiated, the paper could offer a low-cost visual alternative to LiDAR or 3D scanners for coarse log sorting in small operations. The paper has positive features: it uses a public dataset, reports inference speed, is transparent about CPU-only training, and candidly lists limitations. However, the reported quantitative metrics concern object detection alone, not diameter estimation; the diameter labels and predictions are both functions of bounding-box width. As presented, the manuscript does not demonstrate estimation of physical timber diameter, so its contribution is at most an anecdotal feasibility demonstration of YOLOv5 for log detection.

major comments (4)
  1. [3.2, 3.3] The diameter-estimation claim is circular as written. Section 3.2 defines the ground-truth diameter category from bounding-box width ("Thin <30px, Medium 30–60px, Thick >60px"), and Section 3.3 states that diameter estimation is a post-processing step using bounding-box width. Since prediction and label are both computed from the detector's box width, agreement on diameter categories is built into the box-width regression and provides no independent evidence about physical log diameter. The phrase "or visual estimate" in Section 3.2 does not resolve this, because no separate visual-label dataset or evaluation is described.
  2. [3.4, 4] Section 3.4 promises that diameter estimation "was assessed using bin accuracy ... and comparison with manually labeled values," but Section 4 reports no such numbers: there is no bin accuracy, confusion matrix, per-class precision/recall for Thin/Medium/Thick, or comparison against manual or physical diameter measurements. The only support for the diameter claim is the sentence that the estimated ranges "appeared visually consistent across multiple test images." This is a qualitative impression, not an evaluation, and it leaves the title and abstract claim unquantified.
  3. [4] Even the detection claim is supported by a single split with no variance or baselines. The test set has only 21 images and 208 instances, and the reported precision 0.656, recall 0.577, mAP@0.5 0.640 come from one training run with no confidence intervals, repeated-seed statistics, or comparison to any alternative detector or feature-based method. Given the small test size, the statement that the model demonstrates "reliable log detection" is stronger than the evidence warrants.
  4. [Table 1, 5] The bounding-box-width proxy is fundamentally uncalibrated. Table 1 reports object widths from 2 px to 4,608 px across dashcam images from forest, roadside, and trailer scenes, which conflates physical diameter with camera distance, perspective, and occlusion. The authors acknowledge this in Section 5 ("results can be affected by camera angles, overlapping objects, or inconsistent scaling"), but the method and evaluation neither correct for nor quantify this effect. A pixel-width bin can only be interpreted as a diameter category if image scale is known or controlled, and no such information is used.
minor comments (6)
  1. [3.2] The description of annotations as "performed using tools provided by Roboflow Universe, which includes wood recognition and annotated log detection sets" is too vague; please specify the exact dataset project, image split, annotation protocol, and any quality-control steps.
  2. [3.3, 4] The training configuration is inconsistent: Section 3.3 says 50–100 epochs with batch size 8–16 and early stopping, while Section 4 says the model was trained for 50 epochs; please report the exact hyperparameters, the number of runs, and how early stopping was applied.
  3. [Figures 2, 3] The figure captions are too generic; Figure 2 should identify the components of the workflow and Figure 3 should describe what is shown in the sample outputs and whether the displayed boxes are predictions, ground truth, or both.
  4. [5] The paragraph in Section 5 about fungal decay, discoloration, and wood identification is tangential to the detection and diameter-estimation results; please integrate it with the limitations discussion or remove it.
  5. [Throughout] There are several presentation issues that need copyediting, including the header title "A L OW-C OST...", inconsistent reference formatting, and the unclear phrase "220 original + 220 prescaled" in Table 1.
  6. [All] No code, trained weights, or data split are made available; a reproducibility statement or link to the fine-tuned model and evaluation code would strengthen the paper.

Circularity Check

1 steps flagged · score 8.0 of 10

Diameter estimation is self-definitional: Sec. 3.2 defines labels as pixel-width thresholds and Sec. 3.3 predicts from the same bounding-box width, so bin agreement is built in; independent physical validation is absent.

  1. self definitional [Section 3.2 (Data Collection and Annotation) and Section 3.3 (Model Architecture)]
    "Labels are as follows: a diameter category defined by visual estimate or bounding box width (e.g., Thin <30px, Medium 30–60px, Thick >60px). ... Diameter estimation is handled as a post-processing step, using bounding box width to assign logs to diameter categories."

    The ground-truth diameter category is defined by thresholds on bounding-box pixel width, and the prediction is computed from the very same bounding-box width. Therefore, once the detector localizes a box, the assigned category is a deterministic function of the box width; agreement with a width-derived label is guaranteed by construction rather than by measuring physical diameter. The paper's promised independent check in Section 3.4 (bin accuracy compared with manually labeled values) is never reported in Section 4, which gives only detection metrics and a qualitative statement that estimates 'appeared visually consistent.' Under the width-based labeling option, the diameter-estimation result reduces to the detection box width, making the central claim circular.

full rationale

The paper contains a genuinely independent component: YOLOv5 detection is evaluated against manually annotated bounding boxes, yielding mAP@0.5 of 0.640 on a held-out test set, and that detection result is not circular. However, the title claim is diameter estimation, and that claim is definitionally tied to bounding-box width. Section 3.2 defines the diameter categories themselves as pixel-width intervals (Thin <30px, Medium 30–60px, Thick >60px), and Section 3.3 assigns logs to those categories by reading the detector's bounding-box width. Hence the 'predicted' category and the 'ground-truth' category are both functions of the same box width; a well-localized box trivially lands in the correct bin. The promised bin-accuracy evaluation against independent manual values is absent from Section 4, which reports only precision, recall, mAP, and a visual-consistency remark. Table 1 also shows object widths ranging from 2 px to 4,608 px, indicating that pixel width varies with camera distance and scene scale, so pixel width is not a validated proxy for physical log diameter. The circularity is limited to the diameter-estimation portion and does not invalidate the detection benchmark, but because diameter estimation is the paper's central contribution, the overall score is high. Correctness concerns about camera angle, perspective, and occlusion are acknowledged in Section 5 but do not cure the definitional circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim depends on one ad hoc assumption (pixel width maps to diameter), two domain assumptions about annotation quality and distribution shift, and four hand-chosen hyperparameters. No new physical or algorithmic entities are introduced.

free parameters (4)
  • Diameter bin thresholds = 30 px and 60 px
    Chosen by hand in Sec 3.2 to define Thin/Medium/Thick. These thresholds determine the diameter categories and are not derived from any physical or statistical criterion.
  • Image resolution = 416x416
    Resolution chosen for training; affects detection accuracy and bbox widths.
  • Epochs = 50-100 with early stopping
    A range reported in Sec 3.3; no analysis of sensitivity.
  • Batch size = 8-16
    Chosen based on Colab CPU constraints (Sec 3.3).
assumptions (4)
  • domain assumption TimberSeg 1.0 bounding-box annotations are accurate and represent true log locations.
    The model is trained and evaluated against these labels; no labeling quality check is reported (Sec 3.2).
  • ad hoc to paper Bounding-box width is a valid proxy for log diameter category.
    The entire diameter-estimation claim rests on this mapping; the paper acknowledges perspective/occlusion sensitivity in Sec 5 but does not correct for it.
  • domain assumption COCO-pretrained YOLOv5 features transfer to log detection.
    Transfer learning is assumed helpful; no ablation without pretraining (Sec 3.3, refs 58 and 59).
  • domain assumption The 21-image test set represents deployment conditions.
    Test images come from the same source distribution as training; the paper notes lack of environmental diversity in Sec 5.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Low-Cost Machine Learning Approach for Timber Diameter Estimation." pith.science (2026). https://pith.science/paper/Y5QYAUXR

@misc{pith2026250717219,
  author       = {Pith},
  title        = {Pith review of: A Low-Cost Machine Learning Approach for Timber Diameter Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y5QYAUXR}},
  note         = {Machine review of arXiv:2507.17219}
}
read the original abstract

The wood processing industry, particularly in facilities such as sawmills and MDF production lines, requires accurate and efficient identification of species and thickness of the wood. Although traditional methods rely heavily on expert human labor, they are slow, inconsistent, and prone to error, especially when processing large volumes. This study focuses on practical and cost-effective machine learning frameworks that automate the estimation of timber log diameter using standard RGB images captured under real-world working conditions. We employ the YOLOv5 object detection algorithm, fine-tuned on a public dataset (TimberSeg 1.0), to detect individual timber logs and estimate thickness through bounding-box dimensions. Unlike previous methods that require expensive sensors or controlled environments, this model is trained on images taken in typical industrial sheds during timber delivery. Experimental results show that the model achieves a mean Average Precision (mAP@0.5) of 0.64, demonstrating reliable log detection even with modest computing resources. This lightweight, scalable solution holds promise for practical integration into existing workflows, including on-site inventory management and preliminary sorting, particularly in small and medium-sized operations.

Figures

Figures reproduced from arXiv: 2507.17219 by the authors.

Figure 1
Figure 1. A sample image from the TimberSeg 1.0 dataset [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Overview of the proposed timber detection and diameter estimation workflow, using a YOLO-based model. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Experimental results Another consideration is that the current detection model uses only one class (“log”), and diameter is either treated as a separate label or inferred post-hoc. This limits its usefulness in tasks requiring more detailed log profiling (e.g., species, size, or quality). Implementing multi-class detection or a multi-task model (e.g., species and diameter jointly) could increase its practical value.… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

59 extracted references · 56 canonical work pages

  1. [1]

    Advanced timber construction industry: A review of 350 multi-storey timber projects from 2000–2021

    Hana Svatoš-Ražnjevi´c, Luis Orozco, and Achim Menges. Advanced timber construction industry: A review of 350 multi-storey timber projects from 2000–2021. Buildings, 12(4):404, 2022

    work page 2000

  2. [2]

    Review of the charring rates of different timber species

    Julie Liu and Erica C Fischer. Review of the charring rates of different timber species. Fire and Materials, 48(1):3–15, 2024

    work page 2024

  3. [3]

    Mechanical properties and durability of some selected timber species

    M Bellal Hossain and ASM Abdul Awal ASM Abdul Awal. Mechanical properties and durability of some selected timber species. Malaysian Journal of Civil Engineering, 24(1), 2012

    work page 2012

  4. [4]

    Fracture mechanics in timber engineering–strength analyses of components and joints

    Erik Serrano and Per Johan Gustafsson. Fracture mechanics in timber engineering–strength analyses of components and joints. Materials and structures, 40:87–96, 2007

    work page 2007

  5. [5]

    Timber application in construction industry and its promotion

    Leopold Mbereyaho, Samuel Tuyishime, J de Montfort, TK Uwintwali, and Chantal Tumukunde. Timber application in construction industry and its promotion. Mediterranean Journal of Basic and Applied Sciences (MJBAS), 3(3):145–154, 2019

    work page 2019

  6. [6]

    Assessment of old timber members: Importance of wood species identification and direct tensile test information

    Jose Saporiti Machado, Filipe Pereira, and Teresa Quilho. Assessment of old timber members: Importance of wood species identification and direct tensile test information. Construction and Building Materials, 207:651–660, 2019. 6 Hasanzadeh Fard et al

    work page 2019

  7. [7]

    An analysis of timber sections and deep learning for wood species classification

    André R de Geus, Sérgio F da Silva, Alexandre B Gontijo, Flávio O Silva, Marcos A Batista, and Jefferson R Souza. An analysis of timber sections and deep learning for wood species classification. Multimedia Tools and Applications, 79(45):34513–34529, 2020

    work page 2020

  8. [8]

    Walter de Gruyter, 2012

    Thomas Herzog, Julius Natterer, Roland Schweitzer, Michael V olz, and Wolfgang Winter.Timber construction manual. Walter de Gruyter, 2012

    work page 2012

Show all 59 references
  1. [9]

    Forensic timber identification: It’s time to integrate disciplines to combat illegal logging

    Eleanor E Dormontt, Markus Boner, Birgit Braun, Gerhard Breulmann, Bernd Degen, Edgard Espinoza, Shelley Gardner, Phil Guillery, John C Hermanson, Gerald Koch, et al. Forensic timber identification: It’s time to integrate disciplines to combat illegal logging. Biological Conse...

    2015

  2. [10]

    Machine learning on dynamic graphs: a survey on applications

    Sanaz Hasanzadeh Fard. Machine learning on dynamic graphs: a survey on applications. 2023 IEEE Ninth Multimedia Big Data (BigMM), pages 32–39, 2023

    2023

  3. [11]

    A two-dimensional self-coordination mechanism of agents in a minority game

    Sanaz Hasanzadeh Fard and Hadi Tabatabaee Malazi. A two-dimensional self-coordination mechanism of agents in a minority game. In International Conference On Fundamentals Of Software Engineering , pages 20–36. Springer, 2019

    2019

  4. [12]

    Face recognition: A literature survey

    Wenyi Zhao, Rama Chellappa, P Jonathon Phillips, and Azriel Rosenfeld. Face recognition: A literature survey. ACM computing surveys (CSUR), 35(4):399–458, 2003

    2003

  5. [13]

    Temporal link prediction using graph embedding dynamics

    Sanaz Hasanzadeh Fard and Mohammad Ghassemi. Temporal link prediction using graph embedding dynamics. In 2023 IEEE Ninth Multimedia Big Data (BigMM), pages 48–55. IEEE, 2023

    2023

  6. [14]

    Fruit sorting robot based on color and size for an agricultural product packaging system

    Tresna Dewi, Pola Risma, and Yurni Oktarina. Fruit sorting robot based on color and size for an agricultural product packaging system. Bulletin of Electrical Engineering and Informatics, 9(4):1438–1445, 2020

    2020

  7. [15]

    Functionalized cellulose nanocrystals for enhanced wood protection: Synthesis, characterization, and performance

    Azar Haghighi Poshtiri, Sima Sepahvand, Mehdi Jonoobi, Alireza Ashori, Ali Naghi Karimi, Fatemeh Hasanzadeh Fard, Laura Bergamonti, Claudia Graiff, and Sabrina Palanti. Functionalized cellulose nanocrystals for enhanced wood protection: Synthesis, characterization, and perform...

    2024

  8. [16]

    The robustness of structural features in species interaction networks

    Sanaz Hasanzadeh Fard and Emily Dolson. The robustness of structural features in species interaction networks. arXiv preprint arXiv:2502.16778, 2025

    2025 arXiv

  9. [17]

    Computer vision: algorithms and applications

    Richard Szeliski. Computer vision: algorithms and applications. Springer Nature, 2022

    2022

  10. [18]

    Computer vision

    Dana Harry Ballard and Christopher M Brown. Computer vision. Prentice Hall Professional Technical Reference, 1982

    1982

  11. [19]

    Computer vision

    George Stockman and Linda G Shapiro. Computer vision. Prentice Hall PTR, 2001

    2001

  12. [20]

    Deep learning for computer vision: A brief review

    Athanasios V oulodimos, Nikolaos Doulamis, Anastasios Doulamis, and Eftychios Protopapadakis. Deep learning for computer vision: A brief review. Computational intelligence and neuroscience, 2018(1):7068349, 2018

    2018

  13. [21]

    Open set recognition of timber species using deep learning for embedded systems

    Marco Paul E Apolinario, Daniel A Urcia Paredes, and Samuel G Huaman Bustamante. Open set recognition of timber species using deep learning for embedded systems. IEEE Latin America Transactions, 17(12):2005–2012, 2019

    2005

  14. [22]

    Timber identification based on the grain structure using learning algorithms

    P Kumar, Sathish Kumar Kannaiah, et al. Timber identification based on the grain structure using learning algorithms. In 2024 International Conference on Smart Technologies for Sustainable Development Goals (ICSTSDG), pages 1–6. IEEE, 2024

    2024

  15. [23]

    A sawn timber tree species recognition method based on am-sppresnet

    Fenglong Ding, Ying Liu, Zilong Zhuang, and Zhengguang Wang. A sawn timber tree species recognition method based on am-sppresnet. Sensors, 21(11):3699, 2021

    2021

  16. [24]

    Yolov1 to yolov10: A comprehensive review of yolo variants and their application in the agricultural domain

    Mujadded Al Rabbani Alif and Muhammad Hussain. Yolov1 to yolov10: A comprehensive review of yolo variants and their application in the agricultural domain. arXiv preprint arXiv:2406.10139, 2024

    2024 arXiv

  17. [25]

    Ap- plication of machine-learning methods in forest ecology: recent progress and future challenges

    Zelin Liu, Changhui Peng, Timothy Work, Jean-Noel Candau, Annie DesRochers, and Daniel Kneeshaw. Ap- plication of machine-learning methods in forest ecology: recent progress and future challenges. Environmental Reviews, 26(4):339–350, 2018

    2018

  18. [26]

    Comparison of machine learning algorithms for forest parameter estimations and application for forest quality assessments

    Qingxia Zhao, Shichuan Yu, Fei Zhao, Linghong Tian, and Zhong Zhao. Comparison of machine learning algorithms for forest parameter estimations and application for forest quality assessments. Forest Ecology and Management, 434:224–234, 2019

    2019

  19. [27]

    A brief review of machine learning algorithms in forest fires science

    Ramez Alkhatib, Wahib Sahwan, Anas Alkhatieb, and Brigitta Schütt. A brief review of machine learning algorithms in forest fires science. Applied Sciences, 13(14):8275, 2023

    2023

  20. [28]

    Machine learning assisted remote forestry health assessment: a comprehensive state of the art review

    Juan Sebastián Estrada, Andrés Fuentes, Pedro Reszka, and Fernando Auat Cheein. Machine learning assisted remote forestry health assessment: a comprehensive state of the art review. Frontiers in plant science, 14:1139232, 2023. 7 Hasanzadeh Fard et al

    2023

  21. [29]

    Classification of thermally treated wood using machine learning techniques

    Vahid Nasir, Sepideh Nourian, Stavros Avramidis, and Julie Cool. Classification of thermally treated wood using machine learning techniques. Wood Science and Technology, 53:275–288, 2019

    2019

  22. [30]

    Wood anatomy and classification of the myrtales

    Ger JCM Van Vliet and Pieter Baas. Wood anatomy and classification of the myrtales. Annals of the Missouri Botanical Garden, pages 783–800, 1984

    1984

  23. [31]

    Leaf and wood classification framework for terrestrial lidar point clouds.Methods in Ecology and Evolution, 10(5):680–694, 2019

    Matheus B Vicari, Mathias Disney, Phil Wilkes, Andrew Burt, Kim Calders, and William Woodgate. Leaf and wood classification framework for terrestrial lidar point clouds.Methods in Ecology and Evolution, 10(5):680–694, 2019

    2019

  24. [32]

    Comparison of wood classification using machine learning

    Agus Pratondo and Astri Novianty. Comparison of wood classification using machine learning. In 2022 IEEE 10th Conference on Systems, Process & Control (ICSPC), pages 308–312. IEEE, 2022

    2022

  25. [33]

    Deep learning for use in lumber classification tasks

    Junfeng Hu, Wenlong Song, Wei Zhang, Yafeng Zhao, and Alper Yilmaz. Deep learning for use in lumber classification tasks. Wood Science and Technology, 53:505–517, 2019

    2019

  26. [34]

    Cost and productivity impacts of product sorting on conventional ground-based timber harvesting operations

    Randy D Cass, Shawn A Baker, and W Dale Greene. Cost and productivity impacts of product sorting on conventional ground-based timber harvesting operations. Forest Products Journal, 59(11-12):108–114, 2009

    2009

  27. [35]

    Stand, stem and log segregation based on wood properties: a review

    Glen Murphy and Dave Cown. Stand, stem and log segregation based on wood properties: a review. Scandinavian Journal of Forest Research, 30(8):757–770, 2015

    2015

  28. [36]

    Analysis of modern wood processing techniques in timber terminals

    OA Kunickaya, Aleksandr Pomiguev, IN Kruchinin, Tamara Storodubtseva, AM V oronova, Dmitry Levushkin, Vyacheslav Borisov, and Viktor Ivanov. Analysis of modern wood processing techniques in timber terminals. Central European Forestry Journal, 68(1):51–59, 2022

    2022

  29. [37]

    Modeling of the process of collection, sorting and transportation of logging residues at the logging area

    Mihail Zyrjanov, Sergey Medvedev, and Tatiana Rjabova. Modeling of the process of collection, sorting and transportation of logging residues at the logging area. Journal of Applied Engineering Science, 19(1):114–118, 2021

    2021

  30. [38]

    The effect of log sorting strategy on the forecasted lumber value after sawing pine wood

    Piotr Taube, Kazimierz Orłowski, Daniel Chuchała, and Jakub Sandak. The effect of log sorting strategy on the forecasted lumber value after sawing pine wood. Acta Facultatis Xylologiae Zvolen, 62:89–102, 2020

    2020

  31. [39]

    Processing and sorting forest residues: Cost, productivity and managerial impacts

    Anil Raj Kizha and Han-Sup Han. Processing and sorting forest residues: Cost, productivity and managerial impacts. Biomass and Bioenergy, 93:97–106, 2016

    2016

  32. [40]

    Improved wood species identification based on multi-view imagery of the three anatomical planes

    Núbia Rosa da Silva, Victor Deklerck, Jan M Baetens, Jan Van den Bulcke, Maaike De Ridder, Mélissa Rousseau, Odemir Martinez Bruno, Hans Beeckman, Joris Van Acker, Bernard De Baets, et al. Improved wood species identification based on multi-view imagery of the three anatomical...

    2022

  33. [41]

    Multiresolution local binary pattern variants based texture feature extraction techniques for efficient classification of microscopic images of hardwood species

    Arvind R Yadav, Radhey Shyam Anand, ML Dewal, and Sangeeta Gupta. Multiresolution local binary pattern variants based texture feature extraction techniques for efficient classification of microscopic images of hardwood species. Applied Soft Computing, 32:101–112, 2015

    2015

  34. [42]

    Automatic classification of native wood charcoal

    Teruo M Maruyama, LS Oliveira, AS Britto Jr, and Silvana Nisgoski. Automatic classification of native wood charcoal. Ecological Informatics, 46:1–7, 2018

    2018

  35. [43]

    Wood species recognition using glcm and correlation

    R Bremananth, B Nithya, and R Saipriya. Wood species recognition using glcm and correlation. In 2009 International Conference on Advances in Recent Technologies in Communication and Computing, pages 615–619. IEEE, 2009

    2009

  36. [44]

    Wood recognition based on grey-level co-occurrence matrix

    Bi-hui Wang, Hang-jun Wang, and Heng-nian Qi. Wood recognition based on grey-level co-occurrence matrix. In 2010 International Conference on Computer Application and System Modeling (ICCASM 2010), volume 1, pages V1–269. IEEE, 2010

    2010

  37. [45]

    A survey of convolutional neural networks: analysis, applications, and prospects

    Zewen Li, Fan Liu, Wenjie Yang, Shouheng Peng, and Jun Zhou. A survey of convolutional neural networks: analysis, applications, and prospects. IEEE transactions on neural networks and learning systems, 33(12):6999– 7019, 2021

    2021

  38. [46]

    An introduction to convolutional neural networks.arXiv preprint arXiv:1511.08458, 2015

    Keiron O’shea and Ryan Nash. An introduction to convolutional neural networks.arXiv preprint arXiv:1511.08458, 2015

    2015 arXiv

  39. [47]

    Tree log identification using convolutional neural networks

    Eero Holmström, Antti Raatevaara, Jonne Pohjankukka, Heikki Korpunen, and Jori Uusitalo. Tree log identification using convolutional neural networks. Smart Agricultural Technology, 4:100201, 2023

    2023

  40. [48]

    Evaluation of wood species identification using cnn-based networks at different magnifica- tion levels

    Khanh Nguyen-Trong. Evaluation of wood species identification using cnn-based networks at different magnifica- tion levels. International Journal of Advanced Computer Science and Applications, 14(4), 2023

    2023

  41. [49]

    Performance of convolutional neural network (cnn) and performance influencing factors for wood species classification of lepidobalanus growing in korea

    Jong-Ho Kim, Wan-Geun Park, and Nam-Hun Kim. Performance of convolutional neural network (cnn) and performance influencing factors for wood species classification of lepidobalanus growing in korea. Scientific Reports, 14(1):18141, 2024. 8 Hasanzadeh Fard et al

    2024

  42. [50]

    Wood identification based on macroscopic images using deep and transfer learning approaches

    Halime Ergun. Wood identification based on macroscopic images using deep and transfer learning approaches. PeerJ, 12:e17021, 2024

    2024

  43. [51]

    Wood species recognition with small data: A deep learning approach

    Yongke Sun, Qizhao Lin, Xin He, Youjie Zhao, Fei Dai, Jian Qiu, and Yong Cao. Wood species recognition with small data: A deep learning approach. International Journal of Computational Intelligence Systems , 14(1):1451–1460, 2021

    2021

  44. [52]

    Object detection using yolo: Challenges, architectural successors, datasets and applications

    Tausif Diwan, G Anirudh, and Jitendra V Tembhurne. Object detection using yolo: Challenges, architectural successors, datasets and applications. multimedia Tools and Applications, 82(6):9243–9275, 2023

    2023

  45. [53]

    Agricultural object detection with you only look once (yolo) algorithm: A bibliometric and systematic literature review

    Chetan M Badgujar, Alwin Poulose, and Hao Gan. Agricultural object detection with you only look once (yolo) algorithm: A bibliometric and systematic literature review. Computers and Electronics in Agriculture, 223:109090, 2024

    2024

  46. [54]

    Lidar as a tool for assessing timber assortments: A systematic literature review

    Cesar Alvites, Marco Marchetti, Bruno Lasserre, and Giovanni Santopuoli. Lidar as a tool for assessing timber assortments: A systematic literature review. Remote Sensing, 14(18):4466, 2022

    2022

  47. [55]

    Using lidar technology in forestry activities

    Abdullah Emin Akay, Hakan O˘guz, Ismail Rakip Karas, and Kazuhiro Aruga. Using lidar technology in forestry activities. Environmental monitoring and assessment, 151:117–125, 2009

    2009

  48. [56]

    ultralytics/yolov5: v3

    Glenn Jocher, Alex Stoken, Jirka Borovec, Liu Changyu, Adam Hogan, Laurentiu Diaconu, Jake Poznanski, Lijun Yu, Prashant Rai, Russ Ferriday, et al. ultralytics/yolov5: v3. 0. Zenodo, 2020

    2020

  49. [57]

    Instance segmentation for autonomous log grasping in forestry operations

    Jean-Michel Fortin, Olivier Gamache, Vincent Grondin, François Pomerleau, and Philippe Giguere. Instance segmentation for autonomous log grasping in forestry operations. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 6064–6071. IEEE, 2022

    2022

  50. [58]

    Transfer learning

    Lisa Torrey and Jude Shavlik. Transfer learning. In Handbook of research on machine learning applications and trends: algorithms, methods, and techniques, pages 242–264. IGI global, 2010

    2010

  51. [59]

    A survey of transfer learning

    Karl Weiss, Taghi M Khoshgoftaar, and DingDing Wang. A survey of transfer learning. Journal of Big data, 3:1–40, 2016. 9

    2016

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.