Deep Learning for Early Plant Disease Detection: A Review and Deployment Framework for Lightweight CNNs (MobileNetV3) on Edge Devices
Keywords:
Edge computing, MobileNetV3, Model quantization, Plant disease classification, Precision agriculture, TinyMLAbstract
Crop diseases threaten yield and food security, while timely expert diagnosis and reliable internet access remain limited in many farming communities. This narrative review examines lightweight convolutional neural networks (CNNs), with emphasis on MobileNetV3, and the model-compression and deployment practices relevant to early plant disease classification on edge devices. A targeted literature search and comparison of published studies, datasets, backbone specifications, and compression methods were used; numerical results are reported from the cited publications rather than from new experiments. The reviewed evidence shows that MobileNetV3-Small achieved approximately 99.50% test accuracy on the controlled PlantVillage dataset in one published study, retaining that accuracy after post-training quantization while reducing the reported parameter count from about 1.5 million to 0.93 million. Other benchmark studies report very high performance on PlantVillage but lower results on more realistic PlantDoc imagery, illustrating that dataset and evaluation conditions materially affect the apparent ranking of architectures. ImageNet accuracy and computational cost are useful for describing pretrained backbones, but they are not substitutes for crop- and disease-specific evaluation. The review therefore outlines a workflow covering field-aware data splitting, transfer learning, quantization, pruning, distillation and conversion to mobile runtimes, and recommends reporting per-class recall and macro-F1 alongside latency, memory, and energy on the target device. The study concludes that reliable deployment depends not only on compact architectures but also on diverse field data, independent field-level testing, uncertainty-aware predictions, and transparent on-device benchmarking. The main obstacle to field use is not model size but the gap between laboratory and field images. Diverse field data, robust evaluation, and explainability are the priority, enabling reliable offline diagnostics for resource-limited farming communities.
References
S. Savary, L. Willocquet, S. J. Pethybridge, P. Esker, N. McRoberts, and A. Nelson, “The global burden of pathogens and pests on major food crops,” Nature Ecology & Evolution, vol. 3, pp. 430–439, Feb. 2019.
A. Kamilaris and F. X. Prenafeta-Boldú, “Deep learning in agriculture: A survey,” Computers and Electronics in Agriculture, vol. 147, pp. 70–90, Apr. 2018.
S. P. Mohanty, D. P. Hughes, and M. Salathé, “Using deep learning for image-based plant disease detection,” Frontiers in Plant Science, vol. 7, Sep. 2016.
D. P. Hughes and M. Salathé, “An open access repository of images on plant health to enable the development of mobile disease diagnostics,” arXiv, 2015.
K. P. Ferentinos, “Deep learning models for plant disease detection and diagnosis,” Computers and Electronics in Agriculture, vol. 145, pp. 311–318, Feb. 2018.
E. C. Too, L. Yujian, S. Njuki, and L. Yingchun, “A comparative study of fine-tuning deep learning models for plant disease identification,” Computers and Electronics in Agriculture, vol. 161, pp. 272–279, Jun. 2019.
A. G. Howard et al., “MobileNets: Efficient convolutional neural networks for mobile vision applications,” arXiv, Apr. 2017.
M. Sandler, A. Howard, M. Zhu, A. Zhmoginov and L. -C. Chen, “MobileNetV2: Inverted Residuals and Linear Bottlenecks,” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018, pp. 4510–4520.
A. Howard et al., “Searching for MobileNetV3,” 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, Korea (South), 2019, pp. 1314–1324.
J. G. A. Barbedo, “Impact of dataset size and variety on the effectiveness of deep learning and transfer learning for plant disease classification,” Computers and Electronics in Agriculture, vol. 153, pp. 46–53, Oct. 2018.
J. G. A. Barbedo, “Factors influencing the use of deep learning for plant disease recognition,” Biosystems Engineering, vol. 172, pp. 84–91, Aug. 2018.
D. Singh, N. Jain, P. Jain, P. Kayal, S. Kumawat, and N. Batra, “PlantDoc: A dataset for visual plant disease detection,” in CoDS COMAD 2020: Proceedings of the 7th ACM IKDD CoDS and 25th COMAD, Jan. 2020, pp. 249–253.
A. Ramcharan, K. Baranowski, P. McCloskey, B. Ahmed, J. Legg, and D. P. Hughes, “Deep learning for image-based cassava disease detection,” Frontiers in Plant Science, vol. 8, Oct. 2017.
J. Liu and X. Wang, “Plant diseases and pests detection based on deep learning: a review,” Plant Methods, vol. 17, 2021.
J. Hu, L. Shen and G. Sun, “Squeeze-and-Excitation Networks,” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018, pp. 7132–7141.
N. Ma, X. Zhang, H.-T. Zheng, and J. Sun, “ShuffleNet V2: Practical guidelines for efficient CNN architecture design,” in Computer Vision, 2018, pp. 122–138.
M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” in Proceedings of the 36th International Conference on Machine Learning, 2019, pp. 6105–6114.
F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, “SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size,” arXiv, Feb. 2016.
A. T. Khan, S. M. Jensen, A. R. Khan, and S. Li, “Plant disease detection model for edge computing devices,” Frontiers in Plant Science, vol. 14, Dec. 2023.
A. S. Shiyan, I. D. Kozlov, I. R. Baimuratov, and N. A. Zhukova, “Recognizing plants and their diseases: Benchmarks for multiclass and multilabel classification,” Pattern Recognition and Image Analysis, vol. 35, pp. 159–168, Jun. 2025.
B. Jacob et al., “Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference,” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 2018, pp. 2704–2713.
S. Han, H. Mao, and W. J. Dally, “Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding,” arXiv, Oct. 2015.
G. Hinton, O. Vinyals, and J. Dean, “Distilling the knowledge in a neural network,” arXiv, Mar. 2015.
R. David et al., “TensorFlow Lite Micro: Embedded machine learning for TinyML systems,” in Proceedings of Machine Learning and Systems, vol. 3, 2021, pp. 800–811.
Ü. Atila, M. Uçar, K. Akyol, and E. Uçar, “Plant leaf disease classification using EfficientNet deep learning model,” Ecological Informatics, vol. 61, Mar. 2021.
R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh and D. Batra, “Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization,” 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 2017, pp. 618–626.
