Deep Transformer Based Anomaly Detection in Sugarcane Red Rot Disease Using Spectral Leaf Images

Rabia Tehseen; Hamza Haider; Muhammad Chaudhry; Uzma Omer

Authors

Rabia Tehseen University of Central Punjab
Hamza Haider University of Central Punjab, Lahore, Pakistan
Muhammad Chaudhry University of Central Punjab, Lahore
Uzma Omer University of Education, Lahore, Pakistan

Keywords:

Hyperspectral Imaging, Red Rot Disease, Spectral–Spatial Learning, Attention Mechanism

Abstract

The study presents a hyperspectral framework using transformers for the detection of Red Rot disease in sugarcane through the use of 31-band hyperspectral leaf images. A set of 400 hyperspectral images was used, with 200 healthy and 200 diseased sugarcane leaves, all within the visible region (400 to 700 nm) being used to develop the proposed method, which includes the use of MST++ spectral reconstruction, wavelength-aware positional encoding, spectral/spatial feature extraction and transformer-based classification for automated monitoring of crop diseases. The Hybrid model outperforms the Spectral-Spatial Transformer (SST) model in classification. Overall classification accuracy is 83.33% for the Hybrid model, while SST achieves 65%. The classified samples of sugarcane that are affected by Red Rot show improved performance on recall (0.90), F1-score (0.84), and sensitivity (86.67%) with the Hybrid framework, in comparison to reduced false negatives and false positives in the confusion matrix analyses. The SST model achieved a validation accuracy of approximately 70%, while the Hybrid models exhibited consistent behavior and evaluated close to 80% in terms of their validation accuracy, with lower training and validation loss values, implying that they performed better in the aspects of convergence and generalization. The Hybrid framework also showed an increase of nearly 18.33% in its classification accuracy when compared with the SST architecture. The proposed Hybrid framework represents a scalable, explainable, and intelligent approach to precision agriculture, monitoring crop diseases, and managing sugarcane disease in a sustainable manner.

References

Dong Bao, Jun Zhou, “Early detection of sugarcane smut and mosaic diseases via hyperspectral imaging and spectral-spatial attention deep neural networks,” J. Agric. Food Res., vol. 18, p. 101369, 2024, doi: https://doi.org/10.1016/j.jafr.2024.101369.

Xiaohui Huang, Yunfei Zhou, “SS-TMNet: Spatial–Spectral Transformer Network with Multi-Scale Convolution for Hyperspectral Image Classification,” Remote Sens., vol. 15, no. 5, p. 1206, 2023, doi: https://doi.org/10.3390/rs15051206.

Z. Zhong, Y. Li, L. Ma, J. Li, and W. S. Zheng, “Spectral–Spatial Transformer Network for Hyperspectral Image Classification: A Factorized Architecture Search Framework,” IEEE Trans. Geosci. Remote Sens., vol. 60, 2022, doi: 10.1109/TGRS.2021.3115699.

Malithi De Silva, Dane Brown, “Multispectral Plant Disease Detection with Vision Transformer–Convolutional Neural Network Hybrid Approaches,” Sensors, vol. 23, no. 20, p. 8531, 2023, doi: https://doi.org/10.3390/s23208531.

Hanxiang Wang, Tri Hai Nguyen, “PD-TR: End-to-end plant diseases detection using a transformer,” Comput. Electron. Agric., vol. 224, 2024, [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0168169924005143

Anuruk Prommakhot, Jakkrit Onshaunjit, Wichian Ooppakaew, Grianggai Samseemoung, Jakkree Srinonchat, “Hybrid CNN and Transformer-Based Sequential Learning Techniques for Plant Disease Classification,” IEEE Access, 2025, [Online]. Available: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=11072169

Manh Tuan Do, Manh Hung Ha, “Toward improving precision and complexity of transformer-based cost-sensitive learning models for plant disease detection,” Front. Comput. Sci., vol. 6, 2024, doi: https://doi.org/10.3389/fcomp.2024.1480481.

Ali Zia, Jun Zhou, Muyiwa Olayemi, “Determining Mosaic Resilience in Sugarcane Plants using Hyperspectral Images,” arXiv:2501.16700, 2025, [Online]. Available: https://arxiv.org/abs/2501.16700

Aryan Kumar, Gaurav Saini, “A Comparative Study of Deep Learning Approaches for Early Detection of Sugarcane Diseases,” Procedia Comput. Sci., vol. 260, pp. 182–190, 2025, doi: https://doi.org/10.1016/j.procs.2025.03.192.

Lan Zhang, Yang Wang, “A multi-range spectral-spatial transformer for hyperspectral image classification,” Infrared Phys. Technol., vol. 135, p. 104983, 2023, doi: https://doi.org/10.1016/j.infrared.2023.104983.

Fernanda Rodrigues Silva, Mário Lúcio V.de Resende, “Molecular Characterization and Pathogenicity of Colletotrichum falcatum Causing Red Rot on Sugarcane in Southern Florida,” J. Fungi, vol. 10, no. 11, p. 742, 2024, doi: https://doi.org/10.3390/jof10110742.

R. Viswanathan, R. Selvakumar, K. Manivannan, R. Nithyanantham, and K. Kaverinathan, “Behaviour of Soil Borne Inoculum of Colletotrichum falcatum in Causing Red Rot in Sugarcane Varieties with Varying Disease Resistance,” Sugar Tech, vol. 22, no. 3, pp. 485–497, Jun. 2020, doi: 10.1007/S12355-020-00800-7.

R. Viswanathan and R. Selvakumar, “Varietal Breakdown to Red Rot in Sugarcane Revealed by Comparing Two Colletotrichum falcatum Inoculation Methods,” Sugar Tech 2020 226, vol. 22, no. 6, pp. 1063–1075, Jul. 2020, doi: 10.1007/S12355-020-00855-6.

A. Chandra, A. T. Keizerweerd, Y. Que, and M. P. Grisham, “Loop-mediated isothermal amplification (LAMP) based detection of Colletotrichum falcatum causing red rot in sugarcane,” Mol. Biol. Rep., vol. 42, no. 8, pp. 1309–1316, Aug. 2015, doi: 10.1007/S11033-015-3875-9.

K. P. Ferentinos, “Deep learning models for plant disease detection and diagnosis”, [Online]. Available: https://www.sciencedirect.com/science/article/abs/pii/S0168169917311742?via%3Dihub

Edna Chebet Too, Li Yujian, “A comparative study of fine-tuning deep learning models for plant disease identification,” Comput. Electron. Agric., vol. 161, pp. 272–279, 2019, doi: https://doi.org/10.1016/j.compag.2018.03.032.

A. Khan, A. Sohail, U. Zahoora, and A. S. Qureshi, “A survey of the recent architectures of deep convolutional neural networks,” Artif. Intell. Rev. 2020 538, vol. 53, no. 8, pp. 5455–5516, Apr. 2020, doi: 10.1007/S10462-020-09825-6.

D. Wang et al., “Early Detection of Tomato Spotted Wilt Virus by Hyperspectral Imaging and Outlier Removal Auxiliary Classifier Generative Adversarial Nets (OR-AC-GAN),” Sci. Reports 2019 91, vol. 9, no. 1, pp. 1–14, Mar. 2019, doi: 10.1038/s41598-019-40066-y.

Muhammad Shafay, Taimur Hassan, Muhammad Owais, Irfan Hussain, Sajid Gul Khawaja, Lakmal Seneviratne & Naoufel Werghi, “Recent advances in plant disease detection: challenges and opportunities,” Plants Methods, vol. 21, no. 140, 2025, [Online]. Available: https://link.springer.com/article/10.1186/s13007-025-01450-0

I. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, “Attention is all you need,” Adv. Neural Inf. Process. Syst., p. 30, 2017.

J. Hu, L. Shen, S. Albanie, G. Sun, and E. Wu, “Squeeze-and-Excitation Networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 42, no. 8, pp. 2011–2023, Aug. 2020, doi: 10.1109/TPAMI.2019.2913372.

K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., vol. 2016-December, pp. 770–778, Dec. 2016, doi: 10.1109/CVPR.2016.90.

S Aswani, Ashish Sinha, “Early diagnosis of Sugarcane leaf diseases through CNN and Vision Transformer hybrid model,” Researchgate, 2025, doi: 10.21203/rs.3.rs-7232200/v1.