Resolving Visual Ambiguity in Wood Grain Classification  Using InceptionV3 and Label Smoothing

Rianto Rianto; Sulistyo Dwi Sancoko; Eko Setyo Humanika

doi:10.29207/resti.v10i3.7599

Authors

Rianto Rianto Universitas Teknologi Yogyakarta https://orcid.org/0000-0002-5058-4580
Sulistyo Dwi Sancoko Universitas Teknologi Yogyakarta https://orcid.org/0000-0002-5658-5402
Eko Setyo Humanika Universitas Teknologi Yogyakarta https://orcid.org/0000-0001-9789-0764

DOI:

https://doi.org/10.29207/resti.v10i3.7599

Keywords:

fine-grained classification, InceptionV3, label smoothing, visual ambiguity, wood grain patterns

Abstract

Wood type classification in the timber industry is frequently hindered by the high visual similarity of grain patterns, a structural challenge that is particularly pronounced with small-scale datasets. This study systematically resolves this fine-grained visual ambiguity by developing a rigorously optimized InceptionV3 framework. Initial baseline evaluations conducted on a dataset of 800 images encompassing four wood species, namely Pterospermum javanicum (Bayur), Magnolia champaca (Cempaka), Tectona grandis (Jati), and Melia azedarach (Mindi), revealed that a standard ResNet50 architecture experienced a severe performance degradation, generating a random guess accuracy of merely 25 percent. Addressing this severe architectural inadequacy, the proposed InceptionV3 model integrates highly targeted spatial augmentations, a Dropout rate of 0.6, and L2 regularization. Furthermore, the optimization strategy deploys Stochastic Gradient Descent with Nesterov momentum and Label Smoothing to explicitly mitigate intra-class visual similarities. Consequently, the proposed framework achieved a robust overall accuracy of 89 percent, surpassing the baseline by a substantial 64 percent, alongside a macro-averaged F1 Score of 0.89. These empirical findings substantiate that specific architectural fine-tuning and advanced stochastic regularization are highly essential for capturing subtle wood-grain patterns, thereby offering a highly reliable, automated quality-control solution for the timber industry.

Downloads

Download data is not yet available.

References

E. Korkmaz, C. Özcan, and M. Korkmaz, “An application of fuzzy soft sets to a real-life problem: Classification of wood materials to prevent fire-related injuries and deaths,” Appl. Soft Comput., vol. 132, p. 109875, Jan. 2023, doi: 10.1016/J.ASOC.2022.109875.

J. Zhu, H. Peng, X. Lu, J. Lyu, and T. Zhan, “Classification of Principal Wood Species in China Based on the Physiomechanical Properties,” J. Renew. Mater., vol. 11, no. 3, pp. 1425–1437, Oct. 2022, doi: 10.32604/JRM.2022.023464.

W. Gan et al., “Enhanced Recognition of Sustainable Wood Building Materials Based on Deep Learning and Augmentation,” Sustainability 2025, Vol. 17, vol. 17, no. 15, Jul. 2025, doi: 10.3390/su17156683.

F. Fang, Q. Xu, Y. Cheng, Y. Sun, and J. H. Lim, “Image Understanding With Reinforcement Learning: Auto-Tuning Image Attributes and Model Parameters for Object Detection and Segmentation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 10, pp. 6671–6685, Oct. 2022, doi: 10.1109/TCSVT.2022.3171781.

A. Pratondo and A. Novianty, “Comparison of Wood Classification using Machine Learning,” 2022 IEEE 10th Conference on Systems, Process and Control, ICSPC 2022 - Proceedings, pp. 308–312, 2022, doi: 10.1109/ICSPC55597.2022.10001776.

H. Ergun, “Wood identification based on macroscopic images using deep and transfer learning approaches,” PeerJ, vol. 12, p. e17021, Feb. 2024, doi: 10.7717/PEERJ.17021/SUPP-1.

R. Ehtisham et al., “Classification of defects in wooden structures using pre-trained models of convolutional neural network,” Case Studies in Construction Materials, vol. 19, p. e02530, Dec. 2023, doi: 10.1016/J.CSCM.2023.E02530.

S. Salvi and V. Geetha, “A Nested Texture Inspired Novel Image Pattern Based Optical Camera Communication,” IEEE Access, vol. 10, pp. 109056–109067, 2022, doi: 10.1109/ACCESS.2022.3213701.

D. A. Abdulqader, M. S. Hathal, B. M. Mahmmod, S. H. Abdulhussain, and D. Al-Jumeily, “Plain, Edge, and Texture Detection Based on Orthogonal Moment,” IEEE Access, vol. 10, pp. 114455–114468, 2022, doi: 10.1109/ACCESS.2022.3217225.

E. S. Gemeay, F. A. Alenizi, A. H. Mohammed, M. H. Shakoor, and R. Boostani, “Weighted Constraint Feature Selection of Local Descriptor for Texture Image Classification,” IEEE Access, vol. 11, pp. 91673–91695, 2023, doi: 10.1109/ACCESS.2023.3306075.

Z. Tian, N. Zhang, J. Wang, L. Sha, H. Liu, and L. Zou, “Image Classification of Imbalanced Wood Microscope by Integrating Multi-source Features,” International Journal of Computational Intelligence Systems 2025 18:1, vol. 18, no. 1, pp. 1–16, Jul. 2025, doi: 10.1007/S44196-025-00912-6.

D. Mane et al., “A Novel Fuzzy Hypersphere Neural Network Classifier using Class Specific Clustering for Robust Pattern Classification,” IEEE Access, 2024, doi: 10.1109/ACCESS.2024.3454296.

K. M. Zielinski et al., “Advanced wood species identification based on multiple anatomical sections and using deep feature transfer and fusion,” Comput. Electron. Agric., vol. 231, p. 109867, Apr. 2025, doi: 10.1016/J.COMPAG.2024.109867.

Q. Wang, Y. Dong, N. Xu, F. Xu, C. Mou, and F. Chen, “Image Classification of Tree Species in Relatives Based on Dual-Branch Vision Transformer,” Forests 2024, Vol. 15, vol. 15, no. 12, Dec. 2024, doi: 10.3390/f15122243.

Y. Wang, Z. P. Bian, J. Hou, and L. P. Chau, “Convolutional Neural Networks with Dynamic Regularization,” IEEE Trans. Neural Netw. Learn. Syst., vol. 32, no. 5, pp. 2299–2304, May 2021, doi: 10.1109/TNNLS.2020.2997044.

L. P. Yi, M. F. Akbar, M. N. A. Wahab, B. A. Rosdi, M. A. Fauthan, and N. H. M. M. Shrifan, “The Prospect of Artificial Intelligence-Based Wood Surface Inspection: A Review,” IEEE Access, vol. 12, pp. 84706–84725, 2024, doi: 10.1109/ACCESS.2024.3412928.

R. Confalonieri, P. P. Htun, B. Sun, and T. Tillo, “An End-to-End Framework for the Classification of Hyperspectral Images in the Wood Domain,” IEEE Access, vol. 12, pp. 38908–38916, 2024, doi: 10.1109/ACCESS.2024.3376258.

T. H. Wong, M. F. Soong, A. Saifizul, and R. Ramli, “Assessment of wood quality for Sape making: Vibroacoustic analysis and machine learning classification,” Expert Syst. Appl., vol. 261, p. 125591, Feb. 2025, doi: 10.1016/J.ESWA.2024.125591.

İ. Kırbaş and A. Çifci, “An effective and fast solution for classification of wood species: A deep transfer learning approach,” Ecol. Inform., vol. 69, p. 101633, Jul. 2022, doi: 10.1016/J.ECOINF.2022.101633.

T. Hidayat, I. A. Astuti, A. Yaqin, A. P. Tjilen, and T. Arifianto, “Grouping of Image Patterns Using Inceptionv3 For Face Shape Classification,” JOIV : International Journal on Informatics Visualization, vol. 7, no. 4, pp. 2336–2343, Dec. 2023, doi: 10.62527/joiv.7.4.1743.

E. Mengiste, K. R. Mannem, S. A. Prieto, and B. Garcia de Soto, “Transfer-Learning and Texture Features for Recognition of the Conditions of Construction Materials with Small Data Sets,” Journal of Computing in Civil Engineering, vol. 38, no. 1, p. 04023036, Jan. 2024, doi: 10.1061/jccee5.cpeng-5478.

R. K. Ahalya, U. Snekhalatha, and V. Dhanraj, “Automated segmentation and classification of hand thermal images in rheumatoid arthritis using machine learning algorithms: A comparison with quantum machine learning technique,” J. Therm. Biol., vol. 111, p. 103404, Jan. 2023, doi: 10.1016/J.JTHERBIO.2022.103404.

J. Zheng et al., “An improved automatic image labeling and classification algorithm for multi-mode damage quantification of 2.5D woven composites based on deep learning strategy,” Compos. Sci. Technol., vol. 259, p. 110932, Jan. 2025, doi: 10.1016/J.COMPSCITECH.2024.110932.

H. Ren, Y. Zhao, Y. Zhang, and W. Sun, “Learning label smoothing for text classification,” PeerJ Comput. Sci., vol. 10, p. e2005, Apr. 2024, doi: 10.7717/PEERJ-CS.2005.

T. H. Cheung and D. Y. Yeung, “A Survey of Automated Data Augmentation for Image Classification: Learning to Compose, Mix, and Generate,” IEEE Trans. Neural Netw. Learn. Syst., vol. 35, no. 10, pp. 13185–13205, 2024, doi: 10.1109/TNNLS.2023.3282258.

S. Sharma, K. Guleria, S. Kumar, and S. Tiwari, “Deep Learning based Model for Detection of Vitiligo Skin Disease using Pre-trained Inception V3,” Deep Learning based Model for Detection of Vitiligo Skin Disease using Pre-trained Inception V3, vol. 8, no. 5, pp. 1024–1039, Jul. 2023, doi: 10.33889/IJMEMS.2023.8.5.059.

S. Lee, S. Lee, and B. C. Song, “Improving Vision Transformers to Learn Small-Size Dataset From Scratch,” IEEE Access, vol. 10, pp. 123212–123224, 2022, doi: 10.1109/ACCESS.2022.3224044.