Penggunaan Transfer Learning untuk Peningkatan Akurasi Deteksi Penyakit Tanaman
Article Metrics
Abstract view : 675 timesAbstract
Penyakit tanaman menjadi tantangan utama dalam meningkatkan produktivitas pertanian dan ketahanan pangan. Teknologi deep learning, khususnya Convolutional Neural Networks (CNN), telah digunakan secara luas untuk mendeteksi penyakit tanaman melalui citra daun. Namun, efektivitas model CNN sering menurun saat dihadapkan pada data dengan distribusi berbeda dari data pelatihan awal (domain shift). Penelitian ini bertujuan meningkatkan akurasi klasifikasi penyakit tanaman menggunakan pendekatan transfer learning dengan model pralatih VGG16, ResNet50, dan DenseNet121 yang diadaptasi pada dataset PlantVillage. Selain itu, digunakan teknik augmentasi data baik secara manual (rotasi, flipping) maupun sintetis melalui Generative Adversarial Networks (GANs) untuk meningkatkan generalisasi model. Hasil evaluasi menunjukkan bahwa DenseNet121 memberikan performa terbaik dengan akurasi 95,2%, precision 0,94, recall 0,95, dan F1-score 0,94. Penggunaan augmentasi data terbukti meningkatkan kemampuan model dalam menghadapi variasi latar belakang dan pencahayaan, terutama pada dataset eksternal. Temuan ini menunjukkan bahwa kombinasi transfer learning dan augmentasi data mampu menghasilkan model klasifikasi yang lebih robust terhadap perbedaan domain dan lebih adaptif untuk implementasi pada kondisi lapangan yang nyata.
References
Bettelli, M. et al. (2023) ‘Classification and Forecasting of Water Stress in Tomato Plants Using Bioristor Data’, IEEE Access, 11(April), pp. 34795–34807. doi: 10.1109/ACCESS.2023.3265597.
Dash, A., Sethy, P. K. and Behera, S. K. (2023) ‘Maize disease identification based on optimized support vector machine using deep feature of DenseNet201’, Journal of Agriculture and Food Research, 14(August), p. 100824. doi: 10.1016/j.jafr.2023.100824.
Dehbozorgi, P., Ryabchykov, O. and Bocklitz, T. (2024) ‘A Systematic Investigation of Image Pre-Processing on Image Classification’, IEEE Access, 12(February), pp. 64913–64926. doi: 10.1109/ACCESS.2024.3395063.
Deng, H. et al. (2021) ‘Rahc_gan: A data augmentation method for tomato leaf disease recognition’, Symmetry, 13(9). doi: 10.3390/sym13091597.
Dheeraj, A. and Chand, S. (2024) ‘LWDN: lightweight DenseNet model for plant disease diagnosis’, Journal of Plant Diseases and Protection, 131(3), pp. 1043–1059. doi: 10.1007/s41348-024-00915-z.
Jlassi, A. et al. (2024) ‘Potato Leaf Disease Classification Using Transfer Learning and Reweighting-Based Training with Imbalanced Data’, SN Computer Science, 5(8). doi: 10.1007/s42979-024-03334-x.
Kumar, T. et al. (2023) ‘Image Data Augmentation Approaches: A Comprehensive Survey and Future directions’, 12(December). doi: 10.1109/ACCESS.2024.3470122.
Lanjewar, M. G., Morajkar, P. and P, P. (2024) ‘Modified transfer learning frameworks to identify potato leaf diseases’, Multimedia Tools and Applications, 83(17), pp. 50401–50423. doi: 10.1007/s11042-023-17610-0.
Li, S. et al. (2024) ‘Application of Semi-Supervised Learning in Image Classification: Research on Fusion of Labeled and Unlabeled Data’, IEEE Access, 12(February), pp. 27331–27343. doi: 10.1109/ACCESS.2024.3367772.
Liu, H. et al. (2022) ‘TBEM: Testing-Based GPU-Memory Consumption Estimation for Deep Learning’, IEEE Access, 10, pp. 39674–39680. doi: 10.1109/ACCESS.2022.3164510.
Liu, W., Yu, L. and Luo, J. (2022) ‘A hybrid attention-enhanced DenseNet neural network model based on improved U-Net for rice leaf disease identification’, Frontiers in Plant Science, 13(October), pp. 1–16. doi: 10.3389/fpls.2022.922809.
Santoso, H. A. et al. (2024) ‘Comparative analysis of convolutional neural network and DenseNet121 transfer learning in agriculture focusing on crop leaf disease identification’, Applied Computing and Informatics. doi: 10.1108/ACI-03-2024-0132.
Swapno, S. M. M. R. et al. (2025) ‘ViT-SENet-Tom: machine learning-based novel hybrid squeeze–excitation network and vision transformer framework for tomato fruits classification’, Neural Computing and Applications, 5, pp. 7–14. doi: 10.1007/s00521-025-10973-5.
Wu, Q. et al. (2023) ‘A classification method for soybean leaf diseases based on an improved ConvNeXt model’, Scientific Reports, 13(1), pp. 1–11. doi: 10.1038/s41598-023-46492-3.
Zhuang, H. et al. (2024) ‘An Attention-Based Deep Learning Network for Predicting Platinum Resistance in Ovarian Cancer’, IEEE Access, 12(March), pp. 41000–41008. doi: 10.1109/ACCESS.2024.3377560.
Copyright (c) 2025 Tuti Andriani
This work is licensed under a Creative Commons Attribution 4.0 International License.
