Journal of Researches in Mechanics of Agricultural Machinery

Journal of Researches in Mechanics of Agricultural Machinery

A New Hybrid Deep Learning Multi-Domain Framework for Early Detection of Apple Leaf Diseases

Document Type : Special English issue

Authors
Forough Ehsanfar
Fatemeh Moayedi
Larestan Higher Education Center
10.22034/jrmam.2026.14992.763
Abstract
Rapid and accurate detection of plant diseases, particularly in apple orchards, plays a vital role in improving productivity and minimizing economic losses. Many previous studies in this field rely on computationally intensive models that struggle to generalize effectively under real-world and cross-domain conditions. In this study, a lightweight, accurate, and environment-robust model is proposed for apple leaf disease classification. Two pre-trained architectures, ResNet18 and EfficientNet-B0, are employed as the feature extraction backbones, while their classification heads are replaced with a newly designed two-layer MLP-based head. In this design, the extracted features are first passed through a fully connected layer with 256 neurons, followed by a ReLU activation function and a Dropout layer to mitigate overfitting, and finally connected to a fully connected output layer with a neuron count corresponding to the number of disease classes. This structure not only reduces overfitting but also enables learning nonlinear decision boundaries, enhancing model performance under multi-domain scenarios. Experimental results on the 9-class AL9EE dataset demonstrate that the proposed model achieves 98% accuracy while maintaining computational efficiency, striking an optimal balance between accuracy and processing cost. Furthermore, its simplicity and high inference speed make it highly suitable for deployment on edge devices and IoT-based agricultural systems. The findings highlight that integrating deep learning architectures with re-engineered final shallow learning layers provides an effective and practical pathway toward developing intelligent and real-time plant disease detection systems for smart agriculture applications.
Keywords
Apple leaf disease
hybrid deep learning
. multi-domain approach
. EfficientNet-B0
Edge computing
Agricultural Research Center of Iran. (2022). Survey on plant diseases in apple orchards. (In Persian).
 
Ali, H., Shifa, N., Benlamri, R., Farooque, A. A., & Yaqub, R. (2025). A fine tuned EfficientNet-B0 convolutional neural network for accurate and efficient classification of apple leaf diseases. Scientific Reports, 15(1), 25732.
   https://doi.org/10.1038/s41598-025-04479-2
 
Ali, M. U., Khalid, M., Farrash, M., Lahza, H. F. M., Zafar, A., & Kim, S. H. (2024). AppleLeafNet: A lightweight and efficient deep learning framework for diagnosing apple leaf diseases. Frontiers in Plant Science, 15, 1502314. https://doi.org/10.3389/fpls.2024.1502314
 
Anam, S. (2020). Segmentation of leaf spot disease in apple plants using particle swarm optimization and K-means algorithm. Journal of Physics: Conference Series, 1562, 012011. https://doi.org/10.1088/1742-6596/1562/1/012011
 
Ashmafee, M. H., Ahmed, T., Ahmed, S., Hasan, M. B., Jahan, M. N., & Rahman, A. A. (2023, February). An efficient transfer learning-based approach for apple leaf disease classification. In 2023 International Conference on Electrical, Computer and Communication Engineering (ECCE) (pp. 1-6). IEEE. https://doi.org/10.1109/ECCE57851.2023.10101542
 
Birant, D. (2024). PlantVillage dataset. Figshare. https://doi.org/10.6084/m9.figshare.26096800.v1
 
Bracino, A. A., Concepcion, R. S., Bedruz, R. A. R., Dadios, E. P., & Vicerra, R. R. P. (2020, December). Development of a hybrid machine learning model for apple (Malus domestica) health detection and disease classification. In 2020 IEEE 12th international conference on humanoid, nanotechnology, information technology, communication and control, environment, and management (HNICEM) (pp. 1-6). IEEE. https://doi.org/10.1109/HNICEM51456.2020.9400139
 
Chakraborty, S., Paul, S., & Rahat-uz-Zaman, M. (2021). Prediction of apple leaf diseases using multiclass support vector machine. In 2021 2nd International Conference on Robotics, Electrical and Signal Processing Techniques (ICREST) (pp. 1–6). https://doi.org/10.1109/ICREST51555.2021.9331132
 
Food and Agriculture Organization of the United Nations. (2025). FAOSTAT: Crops and livestock products. Retrieved August 30, 2025, from https://www.fao.org/faostat/en/#data
 
Fu, J., Li, S., Sun, Y., Mu, Y., Hu, T., & Gong, H. (2022). Lightweight CNN for apple leaf disease recognition. Frontiers in Plant Science, 13, 831219. https://doi.org/10.3389/fpls.2022.831219
 
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 770–778). IEEE. https://doi.org/10.1109/CVPR.2016.90
 
Howard, J., & Gugger, S. (2020). Deep learning for coders with fastai & PyTorch. O’Reilly Media.
 
Hughes, D., & Salathé, M. (2015). An open access repository of images on plant health to enable the development of mobile disease diagnostics. arXiv preprint arXiv:1511.08060.
 
Khan, M. A., Lali, M. I. U., Sharif, M., Javed, K., Aurangzeb, K., Haider, S. I., ... & Akram, T. (2019). An optimized method for segmentation and classification of apple diseases based on strong correlation and genetic algorithm based feature selection. IEEE Access, 7, 46261-46277.https://doi.org/10.1109/ACCESS.2019.2908040
 
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017). ImageNet classification with deep convolutional neural networks. Communications of the ACM, 60, 84–90. https://doi.org/10.1145/3065386
 
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., ... & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. the Journal of machine Learning research, 12, 2825-2830.Sai, A. M., &
 
Patil, N. (2022). Comparative analysis of machine learning algorithms for disease detection in apple leaves. In 2022 International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER) (pp. 239–244). https://doi.org/10.1109/DISCOVER55800.2022.9974840
 
Szegedy, C., Ioffe, S., Vanhoucke, V., & Alemi, A. (2017, February). Inception-v4, inception-resnet and the impact of residual connections on learning. In Proceedings of the AAAI conference on artificial intelligence (Vol. 31, No. 1).
 
Tan, M., & Le, Q. V. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. In Proceedings of the 36th International Conference on Machine Learning (ICML) (pp. 6105–6114).
 
Zhu, F. (2023). AppleLeaf9-Enhanced Edition (AL9EE) dataset. Figshare. https://doi.org/10.6084/m9.figshare.23606010.v1