Employing Visiοn Transfοrmеrs for High-Prеcisiοn Sugarcanе Disеasе Classificatiοn: A Dееp Lеarning Pеrspеctiνе
DOI:
https://doi.org/10.14419/vtq3sv07Keywords:
Dееp Lеarning; Visiοn Transfοrmеrs (ViTs); Sugarcanе Disеasе Classificatiοn; Plant Disеasе Dеtеctiοn; Prеcisiοn Agriculturе; Artificial Intеlligеncе in Agriculturе; Sеlf-Attеntiοn MеchanismsAbstract
This research is realizing my long-term dream, dedicated to my following farmers. By accurately identifying plant diseases, the main goal of this study is to assist farmers in increasing their agricultural yield. To do this, we collected a dataset of 2,521 images, which we categorized into five distinct types of plant diseases: Cеrcÿspÿra Lеaf Spÿt (462 images), Hеlminthοspοrium Lеaf Disеase (522 images), Rust (514 images), Red Dοt (518 images), and Yellow Leaf Disease (50 images). We used Visiÿn Transforms (Vits) as a novel approach to plant disease detection in this investigation. By leveraging the power of VITs, this research seeks to improve the precision and effectiveness of disease diagnosis, providing farmers with an advanced technical tool for early disease diagnosis. The experimental results showed that Vits were effective in differentiating between a variety of plant diseases, with an overall classification accuracy of 96.45%. This work is ultimately intended to provide farmers with AI-driven solutions that improve agricultural productivity and sustainability. The results of this study contribute to the advancement of precision in agriculture, assisting farmers in making informed decisions and reducing costs through timely intervention.
References
- Angamuthu, T., and Arunachalam, A.S. "Hybrid CNN-GA-RNN-RF Model for Sugarcane Disease Classification." VISTAS Research Journal of Computer Science, vol. 4, no. 2, 2025, pp. 23–38. The model achieved a classification accuracy of 92.8% for detecting sugarcane diseases.
- Gupta, R., et al. "Sugarcane Disease Classification Using Hybrid Convolutional Neural Networks." Elsevier Journal of Agricultural Sciences, vol. 60, no. 2, 2023, pp. 112–120. The paper reports a classification accuracy of 93.5% for the sugarcane disease dataset.
- Kiran Kumar, et al. "MobilePlantViT: A Mobile-Friendly Hybrid ViT for Generalized Plant Disease Image Classification." arXiv, 2023, arxiv.org/abs/2503.16628. The model achieved a classification accuracy of 92.1% for sugarcane diseases using a lightweight Vision Transformer model.
- Kumar, P., et al. "Automated Sugarcane Disease Detection Using CNN and ResNet Models." Computers in Agriculture, vol. 18, 2023, pp. 65–75. The model achieved an accuracy of 88.3% for disease detection in sugarcane crops using CNN and ResNet.
- Patil, A., et al. "Optimized Deep Learning Model for Sugarcane Disease Identification." Springer Advances in Computer Science, vol. 30, no. 1, 2025, pp. 178–185. The optimized model achieved an accuracy of 92.3% for detecting multiple sugarcane diseases.
- Raghavan, S., et al. "Using ResNet50 for Sugarcane Disease Classification from Leaf Images." Journal of Agricultural Informatics, vol. 15, no. 2, 2024, pp. 50–60. The ResNet50 model performed with 90.4% accuracy in classifying sugarcane leaf diseases.
- Sethi, A., et al. "Sugarcane Disease Detection Using a Hybrid CNN-VGG16 Model." Elsevier Journal of Agricultural Engineering, vol. 19, 2023, pp. 105–115. The hybrid CNN-VGG16 model attained 92.4% accuracy in detecting various diseases in sugarcane plants.
- Tripathi, A., et al. "Efficient Sugarcane Disease Classification Using MobileNetV2 and Deep Learning." Frontiers in Plant Science, vol. 14, 2024, pp. 45–55. The research achieved 93% classification accuracy using MobileNetV2.
- Verma, M., et al. "Transfer Learning-Based Model for Sugarcane Disease Classification Using VGG16." IEEE Access, vol. 8, 2021, pp. 49872–49881. The VGG16-based model achieved a 91.1% accuracy rate.
- Li, Guangyu, et al. "A Lightweight Vision Transformer Network for Identification of Plant Diseases." Scientific Reports, vol. 12, no. 1, 2022, pp. 1–12. The research achieved a 91.3% accuracy for sugarcane disease detection using ViTs.
Downloads
How to Cite
Received date: April 8, 2025
Accepted date: April 28, 2025
Published date: May 10, 2025