Issue

Vol. 10, No. 4, November 2025

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Modified U-Net for Leaf Segmentation of Eucalyptus pellita Seedlings in Open Natural Environments

https://doi.org/10.22219/kinetik.v10i4.2349
Tegar Alami
IPB University
Yeni Herdiyeni
Wisnu Ananta Kusuma
Budi Tjahjono
Iskandar Zulkarnaen Siregar

Corresponding Author(s) : Yeni Herdiyeni

yeni.herdiyeni@apps.ipb.ac.id

Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control, Vol. 10, No. 4, November 2025

Abstract

Segmenting plant leaves in natural environments was challenging due to fluctuating lighting, complex backgrounds, and heterogeneous leaf morphology. This study was conducted aiming at addressing the above mentioned issues by developing a modified U-Net architecture for segmenting Eucalyptus pellita seedlings in open nursery settings. The proposed solution introduced a ResNet50 encoder pre-trained on ImageNet, enhanced regularization in the decoder, and a combined loss function comprising Binary Cross-Entropy and Dice Loss to optimize pixel-wise accuracy and shape conformity. A total of 2,181 high-resolution RGB images were collected under three distinct lighting conditions: cloudy, sunny, and scorching. All images were manually annotated, stratified, and augmented with geometric and photometric transformations. Model training employed adaptive learning rates and early stopping strategies. The results showed the highest median segmentation score of 0.867 under cloudy conditions, followed by 0.853 under scorching conditions, and 0.838 under sunny conditions. Statistical testing confirmed significant differences across lighting scenarios. Visual inspection further demonstrated the model’s ability to preserve spatial details and mitigate the impact of shadows, reflections, and cluttered backgrounds. Despite the decline in precision under sunny conditions, segmentation consistency remained high. In conclusion, the developed model successfully addressed key challenges in leaf segmentation under variable outdoor lighting. The findings support its use for robust, high-precision segmentation, offering a foundation for real-time plant health monitoring in nursery-scale applications.

Keywords

Leaf segmentation Modified U-Net ResNet50 Eucalyptus pellita Natural environment Deep learning
Alami, T., Herdiyeni, Y., Kusuma, W. A., Tjahjono, B., & Siregar, I. Z. (2025). Modified U-Net for Leaf Segmentation of Eucalyptus pellita Seedlings in Open Natural Environments. Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control, 10(4). https://doi.org/10.22219/kinetik.v10i4.2349
Download Citation
Endnote/Zotero/Mendeley (RIS)
BibTeX
References
  1. T. Santika et al., “Community forest management in Indonesia: Avoided deforestation in the context of anthropogenic and climate complexities,” Glob. Environ. Chang., vol. 46, no. December 2016, pp. 60–71, 2017, doi: 10.1016/j.gloenvcha.2017.08.002.
  2. F. J. Hutapea, L. Volkova, D. S. Mendham, and C. J. Weston, “Eucalyptus pellita substantially outperforms Acacia mangium in tropical savannah ecosystem of Australia, but strategies are needed to maintain soil nutrients,” For. Ecol. Manage., vol. 562, no. February, p. 121930, 2024, doi: 10.1016/j.foreco.2024.121930.
  3. K. S. Allen, R. W. Harper, A. Bayer, and N. J. Brazee, “A review of nursery production systems and their influence on urban tree survival,” Urban For. Urban Green., vol. 21, pp. 183–191, 2017, doi: 10.1016/j.ufug.2016.12.002.
  4. Q. Ali et al., “Power of plant microbiome: A sustainable approach for agricultural resilience,” Plant Stress, vol. 14, no. November, p. 100681, 2024, doi: 10.1016/j.stress.2024.100681.
  5. Y. Faqir, A. Qayoom, E. Erasmus, M. Schutte-Smith, and H. G. Visser, “A review on the application of advanced soil and plant sensors in the agriculture sector,” Comput. Electron. Agric., vol. 226, no. March, p. 109385, 2024, doi: 10.1016/j.compag.2024.109385.
  6. M. Cándido-Mireles, R. Hernández-Gama, and J. Salas, “Detecting vineyard plants stress in situ using deep learning,” Comput. Electron. Agric., vol. 210, no. April 2022, p. 107837, 2023, doi: 10.1016/j.compag.2023.107837.
  7. A. Bin Rashid, A. K. Kausik, A. Khandoker, and S. N. Siddque, “Integration of Artificial Intelligence and IoT with UAVs for Precision Agriculture,” Hybrid Adv., vol. 10, no. January, p. 100458, 2025, doi: 10.1016/j.hybadv.2025.100458.
  8. Y. Deng et al., “An Effective Image-Based Tomato Leaf Disease Segmentation Method Using MC-UNet,” Plant Phenomics, vol. 5, p. 49, 2023, doi: 10.34133/plantphenomics.0049.
  9. Z. Mohammed Amean, T. Low, and N. Hancock, “Automatic leaf segmentation and overlapping leaf separation using stereo vision,” Array, vol. 12, p. 100099, 2021, doi: 10.1016/j.array.2021.100099.
  10. X. Bai et al., “Vegetation segmentation robust to illumination variations based on clustering and morphology modelling,” Biosyst. Eng., vol. 125, pp. 80–97, 2014, doi: 10.1016/j.biosystemseng.2014.06.015.
  11. V. Singh and A. K. Misra, “Detection of plant leaf diseases using image segmentation and soft computing techniques,” Inf. Process. Agric., vol. 4, no. 1, pp. 41–49, 2017, doi: 10.1016/j.inpa.2016.10.005.
  12. L. Lei, Q. Yang, L. Yang, T. Shen, R. Wang, and C. Fu, Deep learning implementation of image segmentation in agricultural applications: a comprehensive review, vol. 57, no. 6. Springer Netherlands, 2024. doi: 10.1007/s10462-024-10775-6.
  13. A. Deshmane, “Lung Image Segmentation with ResNet50 Encoder and U-Net Decoder Lung Image Segmentation with ResNet50 Encoder and U-Net Decoder,” no. December, 2023, doi: 10.13140/RG.2.2.18360.32003.
  14. M. Yang, M. K. Lim, Y. Qu, X. Li, and D. Ni, “Deep neural networks with L1 and L2 regularization for high dimensional corporate credit risk prediction,” Expert Syst. Appl., vol. 213, no. September 2022, p. 118873, 2023, doi: 10.1016/j.eswa.2022.118873.
  15. J. Huang and R. L. Nowack, “Machine Learning Using U-Net Convolutional Neural Networks for the Imaging of Sparse Seismic Data,” Pure Appl. Geophys., vol. 177, no. 6, pp. 2685–2700, 2020, doi: 10.1007/s00024-019-02412-z.
  16. A. Rodríguez-Fernández, C. Blanco-Alegre, A. M. Vega-Maray, R. M. Valencia-Barrera, T. Molnár, and D. Fernández-González, “Effect of prevailing winds and land use on Alternaria airborne spore load,” J. Environ. Manage., vol. 332, no. November 2022, 2023, doi: 10.1016/j.jenvman.2023.117414.
  17. X. Lin et al., “Self-Supervised Leaf Segmentation under Complex Lighting Conditions,” Pattern Recognit., vol. 135, pp. 0–36, 2023, doi: 10.1016/j.patcog.2022.109021.
  18. N. Navab, J. Hornegger, W. M. Wells, and A. F. Frangi, “Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015: 18th International Conference Munich, Germany, October 5-9, 2015 proceedings, part III,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 9351, no. Cvd, pp. 12–20, 2015, doi: 10.1007/978-3-319-24574-4.
  19. K. A. A. Mahmoud, M. M. Badr, N. A. Elmalhy, R. A. Hamdy, S. Ahmed, and A. A. Mordi, “Transfer learning by fine-tuning pre-trained convolutional neural network architectures for switchgear fault detection using thermal imaging,” Alexandria Eng. J., vol. 103, no. May, pp. 327–342, 2024, doi: 10.1016/j.aej.2024.05.102.
  20. L. Qian et al., “Multi-scale context UNet-like network with redesigned skip connections for medical image segmentation,” Comput. Methods Programs Biomed., vol. 243, no. November 2022, p. 802002, 2024, doi: 10.1016/j.cmpb.2023.107885.
  21. Q. Du Nguyen and H. T. Thai, “Crack segmentation of imbalanced data: The role of loss functions,” Eng. Struct., vol. 297, no. August, p. 116988, 2023, doi: 10.1016/j.engstruct.2023.116988.
  22. A. A. Taha and A. Hanbury, “Metrics for evaluating 3D medical image segmentation: Analysis, selection, and tool,” BMC Med. Imaging, vol. 15, no. 1, 2015, doi: 10.1186/s12880-015-0068-x.
  23. L. Maier-Hein et al., “Metrics reloaded: recommendations for image analysis validation,” Nat. Methods, vol. 21, no. 2, pp. 195–212, 2024, doi: 10.1038/s41592-023-02151-z.
  24. N. Abraham and N. M. Khan, “A novel focal tversky loss function with improved attention u-net for lesion segmentation,” Proc. - Int. Symp. Biomed. Imaging, vol. 2019-April, pp. 683–687, 2019, doi: 10.1109/ISBI.2019.8759329.
  25. Y. Lu, S. Young, H. Wang, and N. Wijewardane, “Robust plant segmentation of color images based on image contrast optimization,” Comput. Electron. Agric., vol. 193, no. January, p. 106711, 2022, doi: 10.1016/j.compag.2022.106711.
  26. M. Sultan Mahmud, Q. U. Zaman, T. J. Esau, G. W. Price, and B. Prithiviraj, “Development of an artificial cloud lighting condition system using machine vision for strawberry powdery mildew disease detection,” Comput. Electron. Agric., vol. 158, no. January, pp. 219–225, 2019, doi: 10.1016/j.compag.2019.02.007.
  27. J. A. Vayssade, G. Jones, C. Gée, and J. N. Paoli, “Pixelwise instance segmentation of leaves in dense foliage,” Comput. Electron. Agric., vol. 195, no. February, 2022, doi: 10.1016/j.compag.2022.106797.
  28. L. F. Tian and D. C. Slaughter, “Environmentally adaptive segmentation algorithm for outdoor image segmentation,” Comput. Electron. Agric., vol. 21, no. 3, pp. 153–168, 1998, doi: 10.1016/S0168-1699(98)00037-4.
  29. I. K. Mayanja, C. H. Diepenbrock, V. Vadez, T. Lei, and B. N. Bailey, “Practical Considerations and Limitations of Using Leaf and Canopy Temperature Measurements as a Stomatal Conductance Proxy: Sensitivity across Environmental Conditions, Scale, and Sample Size,” Plant Phenomics, vol. 6, p. 169, 2024, doi: 10.34133/plantphenomics.0169.
  30. Y. Wang, Y. Zhao, and L. Petzold, “An empirical study on the robustness of the segment anything model (SAM),” Pattern Recognit., vol. 155, no. May 2023, p. 110685, 2024, doi: 10.1016/j.patcog.2024.110685.
  31. J. hua ZHANG, F. tao KONG, J. zhai WU, S. qing HAN, and Z. fen ZHAI, “Automatic image segmentation method for cotton leaves with disease under natural environment,” J. Integr. Agric., vol. 17, no. 8, pp. 1800–1814, 2018, doi: 10.1016/S2095-3119(18)61915-X.
  32. X. Li, Z. Sun, S. Lu, and K. Omasa, “PROSPECULAR: A model for simulating multi-angular spectral properties of leaves by coupling PROSPECT with a specular function,” Remote Sens. Environ., vol. 297, no. August, p. 113754, 2023, doi: 10.1016/j.rse.2023.113754.
  33. M. Yuan, D. Wang, J. Lin, S. Yang, and J. Ning, “SSP-MambaNet: An automated system for detection and counting of missing seedlings in glass greenhouse-grown virus-free strawberry,” Plant Phenomics, vol. 7, no. 2, p. 100043, 2025, doi: 10.1016/j.plaphe.2025.100043.
  34. J. Anderegg, R. Zenkl, A. Walter, A. Hund, and B. A. McDonald, “Combining High-Resolution Imaging, Deep Learning, and Dynamic Modeling to Separate Disease and Senescence in Wheat Canopies,” Plant Phenomics, vol. 5, p. 53, 2023, doi: 10.34133/PLANTPHENOMICS.0053.
Read More
Author biographies is not available.
Download this PDF file
Statistic
Read Counter : 0

Downloads

Download data is not yet available.