An Automated Annotation and Lightweight Detection Framework for Rice Leaf Disease Based on Color Space Analysis and YOLOv11

Ningjiang Han; Jiahui Jiang; Zhengzhe Jia; Haodong Yin; Junjie Yang

doi:10.54097/ebs43980

Authors

Ningjiang Han
Jiahui Jiang
Zhengzhe Jia
Haodong Yin
Junjie Yang

DOI:

https://doi.org/10.54097/ebs43980

Keywords:

Disease detection of rice leaves; automatic labeling; YOLOv11; color space analysis; precision farming; an object detector with low weights.

Abstract

Accurate and timely detection of rice leaf diseases is essential for precision agriculture and food security. However, existing object detection approaches heavily rely on labor-intensive manual bounding box annotations, which limits their scalability and practical deployment. In this paper, we propose an automated annotation and detection framework that combines color space analysis-based lesion localization with the lightweight YOLOv11n model for rice leaf disease detection. Our automated annotation method leverages disease-specific HSV color space characteristics and morphological processing to generate bounding box labels without human intervention, achieving a 98.6% successful annotation rate across 5,932 images. The annotated dataset is then used to train YOLOv11n, which achieves a mean Average Precision (mAP@0.5) of 93.36% and mAP@0.5:0.95 of 77.53% on the test set, with only 2.58M parameters and 1.0ms inference time per image. Experimental results on four rice disease categories (Bacterial Blight, Blast, Brown Spot, andTungro) demonstrate that the proposed framework provides an effective and efficient solution for rice disease detection with minimal annotation cost, making it suitable for real-time deployment on edge devices in agricultural settings.

Downloads

Download data is not yet available.

References

[1] Savary, S., Willocquet, L., Pethybridge, S. J., Esker, P., McRoberts, N., & Nelson, A. (2019). Global impact of pathogens and pests on important food crops. Nature Ecology & Evolution, 3(3), 430–439. https://doi.org/10.1038/s41559-019-0833-6.

[2] Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified real-time object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 779–788).

[3] Liu, J., & Wang, X. (2021). Plant diseases and pests detection on basis of deep learning: Review. Plant Methods, 17, 22. https://doi.org/10.1186/s13007-021-00722-9

[4] Li, L., Zhang, S., & Wang, B. (2021). Plant disease detection and classification with the help of deep learning: A review. IEEE Access, 9, 56683–56698. https://doi.org/10.1109/ACCESS.2021.3072807

[5] Jia, L., Wang, T., Chen, Y., Zang, Y., Li, X., Shi, H., & Gao, L. (2023). MobileNet-CA-YOLO: A new version of the YOLOv7 model that is based on the combination of the MobileNetV3 and attention mechanism to detect rice pests and diseases. Agriculture, 13(7), 1285. https://doi.org/10.3390/agriculture13071285

[6] Mohanty, S. P., Hughes, D. P., & Salathé, M. (2016). Deep learning applied to image based detection of plant diseases. Frontiers in Plant Science, 7, 1419. https://doi.org/10.3389/fpls.2016.01419

[7] Ferentinos, K. P. (2018). The models of deep learning in detecting and diagnosing plants with diseases. Computers and Electronics in Agriculture, 145, 311–318. https://doi.org/10.1016/j.compag.2018.02.016

[8] Fuentes, A., Yoon, S., Kim, S. C., & Park, D. S. (2017). Robust deep-learning-based detector of plant diseases and pests recognition in real time based on tomato plants. Sensors, 17(9), 2022. https://doi.org/10.3390/s17092022

[9] Tang, Z., Lu, J., Chen, Z., Qi, F., & Zhang, L. (2023). Enhanced Pest-YOLO: A real-time pest detector that works on the basis of the effective channel attention mechanism and transformer encoder. Ecological Informatics, 78, 102340. https://doi.org/10.1016/j.ecoinf.2023.102340

[10] Ni, H., Shi, Z., Karungaru, S., Lv, S., Li, X., Wang, X., & Zhang, J. (2024). The classification of major pests and diseases of rice based on the ECA attention mechanism. Agriculture, 14(5), 682. https://doi.org/10.3390/agriculture14050682

[11] Sohn, K., Berthelot, D., Carlini, N., Zhang, Z., Zhang, H., Raffel, C. A., & Li, C. L. (2020). FixMatch: Semi-supervised learning made simple through consistency and confidence. In Advances in Neural Information Processing Systems (Vol. 33, pp. 596–608).

[12] Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., & Gustafson, L. (2023). Segment anything. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 4015–4026).

[13] Liu, S., Zeng, Z., Ren, T., Li, F., Zhang, H., & Yang, J. (2024). Grounding DINO: Marrying DINO with grounded pre-training to detect open-set objects. In Proceedings of the European Conference on Computer Vision (pp. 38–55).

[14] Barbedo, J. G. A. (2019). The identification of plant diseases based on individual lesions and spots with the help of deep learning. Biosystems Engineering, 180, 96–107. https://doi.org/10.1016/j.biosystemseng.2019.02.004

[15] Sethy, P. K., Barpanda, N. K., Rath, A. K., & Behera, S. K. (2020). Support vector machine based identification of diseases in rice leaves by deep features. Computers and Electronics in Agriculture, 175, 105527. https://doi.org/10.1016/j.compag.2020.105527

[16] Wang, C. Y., Bochkovskiy, A., & Liao, H. Y. M. (2023). YOLOv7: Trainable bag-of-freebies becomes a new state-of-the-art in real-time object detector. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 7464–7475). https://doi.org/10.1109/CVPR52729.2023.00719