Internal wood defects pose a serious threat to forest health, reduce timber quality, and increase resource waste during processing, thereby constraining the sustainable development of the wood industry. Computed tomography (CT) enables non-destructive, high-resolution visualization of internal structures, yet accurate defect identification and quantitative characterization remain challenging. This study proposed an integrated framework for internal defect detection and 3D reconstruction in industrial wood applications. The proposed SCV-YOLOv8n model, built upon the lightweight YOLOv8n‑seg architecture, integrated a Spatial Pyramid Pooling Faster Cross Stage Partial Channel (SPPFCSPC) module and a Convolutional Block Attention Module (CBAM) to enhance feature extraction and improve segmentation robustness for complex defect textures. Furthermore, an improved Visualization Toolkit (VTK) pipeline was developed for high-precision 3D reconstruction. The upgraded VTK pipeline integrated Backbone-Enhanced Segmentation Refinement (BESR), Feature Connectivity Adjustment (FCA), and Boundary Curvature Guidance (BCG) submodules, which collectively refine boundary continuity, strengthen internal coherence, and preserve curvature consistency, resulting in more faithful representations of internal geometries. Experiments conducted on three typical internal defect types (decay, hollow, and knots) demonstrated that SCV‑YOLOv8n outperformed the baseline YOLOv8n-seg in both detection accuracy and segmentation consistency, while the improved VTK pipeline achieved higher 3D reconstruction fidelity and enhanced morphological realism. The 3D reconstruction results revealed that decay and hollow defects form interpenetrating cylindrical structures, whereas knots present irregular localized volumes. This framework provides an effective and practical method for accurate defect visualization, sawing path optimization, and resource utilization, supporting sustainable and intelligent wood processing.
| [1] |
Alani AM, Giannakis I, Zou L, Lantini L, Tosti F. Reverse-time migration for evaluating the internal structure of tree trunks using ground-penetrating radar. NDT E Int. 2020, 115. 102294
|
| [2] |
Barua S, Lehtonen P, Pahkasalo T. Plantation vision: potentials, challenges and policy options for global industrial forest plantation development. Int for Rev. 2014, 16: 117-127.
|
| [3] |
Bergstra J, Bengio Y. Random search for hyper-parameter optimization. J Mach Learn Res. 2012, 13: 281-305.
|
| [4] |
Boscaini D, Poiesi F, Messelodi S, Younes A, Grande DA. Localisation of defects in volumetric computed tomography scans of valuable wood logs. Pattern Recogn. 2021.
|
| [5] |
Chen Z, Liu H, Meng X, Yue Y, Zhang B, Zou L, Du Y. Internal decay inspection of tree trunks using 3D point cloud and reverse time migration of ground penetrating radar data. NDT E Int. 2023, 137. 102853
|
| [6] |
Chen S, Li Y, Zhang Y, Yang Y, Zhang X. Soft X-ray image recognition and classification of maize seed cracks based on image enhancement and optimized YOLOv8 model. Comput Electron Agric. 2024, 216. 108475
|
| [7] |
Chen Z, Li Z, Zhang S, Fang L, Jiang Q, Zhao F, Zhou B, Zhao H. AutoAlign: pixel-instance feature aggregation for multimodal 3D object detection. Proc Int Joint Conf Artif Intell.. 2022.
|
| [8] |
Coupry B, Brument B, Laurent A, Mélou J, Quéau Y, Durou J. Assessing the quality of 3D reconstruction in the absence of ground truth: application to a multimodal archaeological dataset. Proc IEEE/CVF Winter Conf Appl Comput vis. 2025.
|
| [9] |
Du X, Feng H, Hu M, Fang Y, Chen S. Three-dimensional stress wave imaging of wood internal defects using TKriging method. Comput Electron Agric. 2018, 148: 63-71.
|
| [10] |
Ehtisham R, Qayyum W, Camp C, Plevris V, Mir J, Khan Q, Ahmad A. Computing the characteristics of defects in wooden structures using image processing and CNN. Autom Constr. 2024, 158. 105211
|
| [11] |
FAO. Global forest resources assessment 2020: main report. Rome. 2020.
|
| [12] |
Fukasawa Y. Fungal succession and decomposition of Pinus densiflora snags. Ecol Res. 2018, 33: 435-444.
|
| [13] |
Funt BV, Bryant EC. Detection of internal log defects by automatic interpretation of computer tomography images. For Prod J. 1987, 37: 56-62
|
| [14] |
Gazo R, Vanek J, Abdul-Massih M, Benes B. A fast pith detection for computed tomography scanned hardwood logs. Comput Electron Agric. 2020, 170. 105107
|
| [15] |
Hanwell MD, Martin KM, Chaudhary A, Avila L. The visualization toolkit (VTK): rewriting the rendering code for modern graphics cards. SoftwX. 2015, 1: 9-12.
|
| [16] |
He K, Gkioxari G, Dollár P, Girshick R. Mask R-CNN. Proc IEEE Int Conf Comput vis.. 2017.
|
| [17] |
Hu J, Shen L, Sun G. Squeeze-and-excitation networks. Proc IEEE Conf Comput vis Pattern Recogn. 2018.
|
| [18] |
Jocher G, Chaurasia A, Qiu J (2023) YOLO by ultralytics (version 8.0.0). https://github.com/ultralytics/ultralytics
|
| [19] |
Johansson E, Johansson D, Skog J, Fredriksson M. Automated knot detection for high-speed computed tomography on Pinus sylvestris L. and Picea abies (L.) Karst. using ellipse fitting in concentric surfaces. Comput Electron Agric. 2013, 96: 238-245.
|
| [20] |
Khazem S, Richard A, Fix J, Pradalier C. Deep learning for the detection of semantic features in tree X-ray CT scans. Artif Intell Agric. 2023, 7: 13-26.
|
| [21] |
Krähenbühl A, Kerautret B, Debled-Rennesson I, Mothe F, Longuetaud F. Knot segmentation in 3D CT images of wet wood. Pattern Recogn. 2014, 47: 3852-3869.
|
| [22] |
Lin G, Milan A, Shen C, Reid I. RefineNet: multi-path refinement networks for high-resolution semantic segmentation. Proc IEEE Conf Comput vis Pattern Recogn. 2017.
|
| [23] |
Longuetaud F, Mothe F, Kerautret B, Krähenbühl A, Hory L, Leban J, Debled-Rennesson I. Automatic knot detection and measurements from X-ray CT images of wood: a review and validation of an improved algorithm on softwood samples. Comput Electron Agric. 2012, 85: 77-89.
|
| [24] |
Paszke A, Chaurasia A, Kim S, Culurciello E (2016) ENet: a deep neural network architecture for real-time semantic segmentation. arXiv:1606.02147. https://doi.org/10.48550/arXiv.1606.02147
|
| [25] |
Perlin L, Pinto R, Valle A. Ultrasonic tomography in wood with anisotropy consideration. Constr Build Mater. 2019, 229. 116958
|
| [26] |
Raposo P, Andrade M, Correia J, Salavessa M, Reis C, Oliveira C, Jesus A. Non-destructive structural wood diagnosis of a medieval building. Procedia Struct Integr. 2017, 5: 1147-1152.
|
| [27] |
Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation. Med Image Comput Comput Assist Interv.. 2015.
|
| [28] |
Roussel J, Mothe F, Krähenbühl A, Kerautret B, Debled-Rennesson I, Longuetaud F. Automatic knot segmentation in CT images of wet softwood logs using a tangential approach. Comput Electron Agric. 2014, 104: 46-56.
|
| [29] |
Rummukainen H, Makkonen M, Uusitalo J. Economic value of optical and X-ray CT scanning in bucking of Scots pine. Wood Mater Sci Eng. 2019, 16: 178-187.
|
| [30] |
Sarigul E, Abbott AL, Schmoldt DL. Rule-driven defect detection in CT images of hardwood logs. Comput Electron Agric. 2003, 41: 101-119.
|
| [31] |
Schmoldt DL, He J, Abbott AL. Automated labeling of log features in CT imagery of multiple hardwood species. Wood Fiber Sci. 2000, 32: 287-300
|
| [32] |
Shi J, Li Z, Zhu T, Wang D, Ni C. Defect detection of industry wood veneer based on NAS and multi-channel mask R-CNN. Sens. 2020, 20: 4398.
|
| [33] |
Solimani F, Cardellicchio A, Dimauro G, Petrozza A, Summerer S, Cellini F, Reno V. Optimizing tomato plant phenotyping detection: boosting YOLOv8 architecture to tackle data complexity. Comput Electron Agric. 2024, 218. 108728
|
| [34] |
Sullivan C, Kaszynski A. PyVista: 3D plotting and mesh analysis through a streamlined interface for the Visualization Toolkit (VTK). J Open Source Softw. 2019, 4. 1450
|
| [35] |
Taha AA, Hanbury A. Metrics for evaluating 3D medical image segmentation: analysis, selection, and tool. BMC Med Imaging. 2015, 15. 29
|
| [36] |
Taubin G (1995) A signal processing approach to fair surface design. Proc ACM SIGGRAPH. 351–358
|
| [37] |
Terven J, Córdova-Esparza D, Romero-González J. A comprehensive review of YOLO architectures in computer vision: from YOLOv1 to YOLOv8 and YOLO-NAS. Mach Learn Knowl Extr. 2023, 51680-1716.
|
| [38] |
Varduhi Y, Irina DV. Family of boundary overlap metrics for the evaluation of medical image segmentation. J Med Imaging. 2018, 5. 015006
|
| [39] |
Wang A, Qian W, Li A, Xu Y, Hu J, Xie Y, Zhang L. NVW-YOLOv8s: an improved YOLOv8s network for real-time detection and segmentation of tomato fruits at different ripeness stages. Comput Electron Agric. 2024, 219. 108833
|
| [40] |
Wang Q, Wu B, Zhu P, Li P, Zuo W, Hu Q. ECA-Net: efficient channel attention for deep convolutional neural networks. Proc IEEE/CVF Conf Comput vis Pattern Recogn.. 2020.
|
| [41] |
Xie G, Wang L, Williams RA, Li Y, Zhang P, Gu S. Segmentation of wood CT images for internal defects detection based on CNN: a comparative study. Comput Electron Agric. 2024, 224. 109244
|
| [42] |
Yang S, Wang W, Gao S, Deng P. Strawberry ripeness detection based on YOLOv8 algorithm fused with LW-Swin Transformer. Comput Electron Agric. 2023, 215. 108360
|
| [43] |
Yue X, Wang L, Wacker J, Zhu Z. Electric resistance tomography and stress wave tomography for decay detection in trees: a comparison study. PeerJ. 2019, 7. e6444
|
| [44] |
Zhang X, Sun H, Xu G, Duan Y, den Van Bulcke J, Van Acker J, Shi J. Understanding the effect of knots on mechanical properties of Chinese fir under bending test by using X-ray computed tomography and digital image correlation. For. 2024, 15. 174
|
| [45] |
Zheng Y, Wang M, Zhang B, Shi X, Chang Q. GBCD-YOLO: a high-precision and real-time lightweight model for wood defect detection. IEEE Access. 2024, 12: 12853-12868.
|
RIGHTS & PERMISSIONS
Northeast Forestry University
PDF
