An Interpretable, UAV Imagery-Based Deep Learning Approach for Plant Disease Detection and Severity Estimation in Indian Agriculture | IJCT Volume 13 – Issue 3 | IJCT-V13I3P97
An Interpretable, UAV Imagery-Based Deep Learning Approach for Plant Disease Detection and Severity Estimation in Indian Agriculture | IJCT Volume 13 – Issue 3 | IJCT-V13I3P97
In the vast and diverse agricultural landscape of India, plant diseases remain a persistent and escalating threat to both national food security and the economic stability of millions of farming households. Traditional methods of disease monitoring, which rely heavily on manual field surveys, are increasingly becoming a bottleneck due to chronic labor shortages and the sheer geographic scale of farming operations. To address these systemic inefficiencies, we propose a scalable, deep learning-based solution that leverages Unmanned Aerial Vehicle (UAV) imagery for automated disease detection. Utilizing the MH-SoyaHealthVision dataset, specifically tailored to Indian soybean crops, we have completed the critical first phase of data curation and balancing to address severe class imbalances. Our proposed system involves a robust five-phase architecture designed to identify critical diseases such as Mosaic and Rust with high precision. Crucially, our approach addresses the “trust gap” in AI adoption by moving beyond simple predictions; we integrate Explainable AI (XAI) mechanisms to generate visual heat maps. This report synthesizes findings from recent pivotal studies, details the completed data analysis phase, and outlines the architectural roadmap designed to provide farmers with actionable, interpretable insights for proactive crop management.
Keywords
Deep Learning, UAV, Precision Agriculture, Plant Disease Detection, Explainable AI (XAI), Data Balancing, Indian Agriculture.
Conclusion
Modernizing Indian agriculture requires tools that are not only technologically advanced but also practically accessible and trustworthy. This project addresses the critical gap in localized AI solutions by leveraging the MH-SoyaHealthVision dataset. We have successfully completed the first phase of our roadmap: identifying the research gap and curating a balanced, localized dataset.
By designing a system that goes beyond simple detection to offer interpretability (via XAI) and severity estimation, we aim to provide farmers with a transparent, useful tool. The integration of interpretable heatmaps solves the “black box” problem, fostering trust among users. Moving forward, our focus will shift to the rigorous training of the CNN backbone and fine-tuning the severity estimation modules to ensure high accuracy in real-world field conditions.
References
[1]A. K. Samha et al., “Advanced Crop Health Monitoring for Smart Agriculture Using Fusion of Transfer Learning With Attentive Convolutional Recurrent Neural Network on Remote Sensing Images,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 2025.
[2]M. Kumar R and B. Murugan MS., “Artificial Intelligence based drone for early disease detection and precision pesticide management in cashew farming,” arXiv preprint arXiv:2303.08556, 2023.
[3]S. Shinde and V. N. Attar, “An Indian UAV and leaf image dataset for integrated crop health assessment of soybean crop,” Data in Brief, vol. 50, Art. no. 109517, 2023.
[4]Q. Pan, M. Gao, P. Wu, J. Yan, and S. Li, “A Deep-Learning-Based Approach for Wheat Yellow Rust Disease Recognition from Unmanned Aerial Vehicle Images,” Sensors, vol. 21, no. 19, 2021.
[5]Y. Shi, L. Han, A. Kleerekoper, S. Chang, and T. Hu, “A Novel CropDocNet for Automated Potato Late Blight Disease Detection from the Unmanned Aerial Vehicle-based Hyperspectral Imagery,” arXiv preprint arXiv:2107.13277, 2021.
[6]S. N. A. M. Robi, N. Ahmad, M. A. M. Izhar, H. M. Kaidi, and N. M. Noor, “Utilizing UAV Data for Neural Network-based Classification of Melon Leaf Diseases in Smart Agriculture,” Int. J. Adv. Comput. Sci. Appl., vol. 15, no. 1, 2024.
[7]M. Gunder, et al., “SugarViT-Multi-Objective Regression of UAV Images with Vision Transformers and Deep Label Distribution Learning Demonstrated on Disease Severity Prediction in Sugar Beet,” arXiv preprint arXiv:2311.03076, 2024.
[8]Y. Liu, Y. Song, P. Cui, Y. Fang, and B. Su, “Diagnosis of grapevine leafroll disease severity infection via UAV remote sensing and deep learning,” Smart Agriculture, vol. 5, no. 3, pp. 49-61, 2023.
[9]N. Ye, et al., “Early detection of Citrus Huanglongbing by UAV remote sensing based on MGA-UNET,” Frontiers in Plant Science, vol. 14, Art. no. 1303645, 2023.
[10]
F. T. J. Faria, et al., “PotatoGANs: Utilizing Generative Adversarial Networks, Instance Segmentation, and Explainable AI for Enhanced Potato Disease Identification and Classification,” arXiv preprint arXiv:2405.07332, 2024.
[11]S. P. Mohanty, D. P. Hughes, and M. Salathe, “Using deep learning for´ image-based plant disease detection,” Frontiers in Plant Science, vol. 7,
p. 1419, 2016.
[12]E. C. Tetila et al., “Detection and classification of soybean pests using deep learning with UAV images,” Computers and Electronics in Agriculture, vol. 179, p. 105836, 2020.
[13]J. G. A. Barbedo, “Plant disease severity estimation using optical imaging,” Remote Sensing, vol. 11, no. 11, p. 1376, 2019.
[14]A. Kamilaris and F. X. Prenafeta-Boldu, “Deep learning in agriculture: A´ survey,” Computers and Electronics in Agriculture, vol. 147, pp. 70-90, 2018.
[15]X. Zhang et al., “Swin Transformer for plant disease classification,” Agriculture, vol. 12, no. 8, p. 1123, 2022.
How to Cite This Paper
Gunde Veeraswami, Kadurla Nagaraju, Kamatam Hruthikesh, Mrs.M .Vijaya lakshmi (2026). An Interpretable, UAV Imagery-Based Deep Learning Approach for Plant Disease Detection and Severity Estimation in Indian Agriculture. International Journal of Computer Techniques, 13(3). ISSN: 2394-2231.
