INTEGRASI YOLOv8 DAN AST-DSRA UNTUK DETEKSI DAN PEMODELAN PENYEBARAN GANODERMA KELAPA SAWIT

Abstract

Basal Stem Rot (BSR) caused by Ganoderma boninense remains a major constraint in oil palm plantations due to its soil-borne nature, latent infection phase, and spatial transmission through root system interactions. Although deep learning-based approaches have demonstrated promising performance in detecting Ganoderma from UAV imagery, most existing studies focus primarily on visual classification without considering disease spread dynamics. This study proposes a threshold-validated digital twin approach by integrating YOLOv8 with the Adaptive Spatial-Temporal Disease Spread Risk Assessment (AST-DSRA) framework to detect infected palms while simultaneously modeling spatial transmission risk. An RGB image dataset was independently acquired using a UAV at an altitude of 30 m above the canopy level in an oil palm plantation located in Kubu Raya Regency, West Kalimantan, Indonesia. The dataset was split into training and testing sets at an 80%:20% ratio. YOLOv8 was employed to classify healthy and Ganoderma-infected palms, with model evaluation yielding precision of 89.9%, recall of 90.5%, F1-score of 90.2%, and mAP of 91.4%. AST-DSRA modeled neighborhood-level transmission risk based on inter-tree proximity using a 10 m distance threshold aligned with standard plantation spacing. The results indicate that this integration enables the identification of spatial infection hotspots that cannot be inferred from visual detection alone. The main contribution of this study is the first integration of YOLOv8-based Ganoderma detection and AST-DSRA-based spatial spread analysis within an interpretable and operational digital twin framework for large-scale oil palm disease monitoring.

Keywords: basal stem rot, Ganoderma boninense, oil palm, spatial disease spread, YOLOv8

REFERENSI

[1]      Anwari, A. C. 2024. Klasifikasi penyakit tanaman tomat melalui citra daun menggunakan metode convolutional neural network. Jurnal Ilmu Multidisiplin Universitas, 2(3), 639-647. https://doi.org/10.70294/jimu.v2i03.421

[2]      Atika, E. P., & Maulana, A. S. 2025. Classification of oil palm leaf diseases using YOLOv8-Nano algorithm. Internet of Things and Artificial Intelligence Journal, 5(3), 603-611. https://doi.org/10.31763/iota.v5i3.988

[3]      Escriba-Gelonch, M., Liang, S., & Hessel, V. 2024. Digital twins in agriculture: Orchestration and applications. Journal of Agricultural and Food Chemistry, 72(19), 10737-10752. https://doi.org/10.1021/acs.jafc.4c01934

[4]      Evizal, R., & Prasmatiwi, F. E. 2022. Penyakit busuk pangkal batang dan performa produktivitas kelapa sawit. Jurnal Agrotropika, 21(1), 47. https://doi.org/10.23960/ja.v21i1.5617

[5]      Flood, J., Bridge, P. D., & Pilotti, C. A. 2022. Basal stem rot of oil palm revisited. Annals of Applied Biology, 181, 160-181. https://doi.org/10.1111/aab.12772

[6]      Hutagaol, P., Wirianata, H., & Kristalisasi, E. N. 2024. Perkembangan penyakit busuk batang pada kelapa sawit pasca peremajaan tanaman. Jurnal Agro Industri Perkebunan, 12(1), 41-50. https://doi.org/10.25181/jaip.v12i1.3272

[7]      Hwang, S. F., Strelkov, S. E., Feng, J., Gossen, B. D., Howard, R. J., & Turnbull, G. D. 2011. Influence of cultivar resistance and inoculum density on root hair infection of canola (Brassica napus) by Plasmodiophora brassicae. Plant Pathology, 60(5), 820-829. https://doi.org/10.1111/j.1365-3059.2011.02457.x

[8]      Iliuta, M. E., Pop, E., Caramihai, S. I., & Moisescu, M. A. 2024. Digital twin - A review of the evolution from concept to technology and its analytical perspectives on applications in various fields. Applied Sciences, 14(13), 5454. https://doi.org/10.3390/app14135454

[9]      Lestandy, M., & Nugraha, A. 2025. UAV image classification of oil palm plants using CNN ensemble model. Journal of Applied Informatics, 9(4), 1312-1318.

[10]   Mulya, Y., & Siregar, R. 2024. Rancang bangun sistem pendeteksian hama dan penyakit pada tanaman sawit dengan sensor IoT. Jurnal Teknologi Informasi dan Ilmu Komputer (JTIIK), 11(2), 287-296. https://doi.org/10.25126/jtiik.20241128561

[11]   Mulyatni, A. S., Pratiwi, D., Sasmita, M. A., Darmono, T. W., & Purwantara, A. 2022. Potensi electronic nose untuk mendeteksi penyakit busuk pangkal batang pada kelapa sawit. E-Journal Menara Perkebunan, 90(1). https://doi.org/10.22302/iribb.jur.mp.v90i1.472

[12]   Sanchez-Vega, J. A., Torres-Madronero, M. C., Lopez-Morales, J. A., & Restrepo-Giron, A. D. 2025. Automatic detection of Ceroxylon palms by deep learning in a protected area in Amazonas (NW Peru). Forests, 16(7), 1061. https://doi.org/10.3390/f16071061

[13]   Shehu, H. A., Uyaver, S., & Rasheed, J. 2025. YOLO for early detection and management of Tuta absoluta-induced tomato leaf diseases. Frontiers in Plant Science, 16, 1524630. https://doi.org/10.3389/fpls.2025.1524630

[14]   Zakaria, L. 2023. Basal stem rot of oil palm: The pathogen, disease incidence, and control methods. In S. Sundram & A. Idris (Eds.), Ganoderma Diseases of Perennial Crops: Biology, Detection and Management (pp. 603-615). CABI Publishing. https://doi.org/10.1079/9781800621596.0036

[15]   Zhang, J., Li, Y., Wang, Q., Liu, H., Zhang, J., & Wang, Y. 2024. HR-YOLOv8: A crop growth status object detection method based on YOLOv8. Electronics, 13(9), 1620. https://doi.org/10.3390/electronics13091620

[16]   Zhou, Q., Li, Y., Gu, J., & Shao, J. 2024. Efficient optimized YOLOv8 model with extended vision. Sensors, 24(20), 6506. https://doi.org/10.3390/s24206506