Optimizing Leafnet for Better Rice Leaf Disease Classification

Abstract: Timely identification of plant diseases is crucial for safeguarding crop yields and ensuring effective disease management. In this research, we perform a comparative study of multiple deep learning models for classifying rice leaf diseases using the Rice Leaf Dataset. The classification phase involves evaluating LeafNet, a Modified LeafNet, MobileNetV2, Xception, NasNetMobile, and an ensemble model combining LeafNet with NasNetMobile. For leaf abnormality detection, we employ object detection architectures from the YOLO family, including YOLOV5x6, YOLOV5s6, YOLOV8n, and YOLOV9n. The primary objective is to enhance accuracy in both disease classification and abnormality detection, thereby supporting precision agriculture practices. Through extensive experimentation, we identify the models that deliver superior performance in their respective tasks. The findings highlight the value of advanced machine learning approaches in modern agriculture, enabling early intervention, efficient disease control, and optimized resource use in rice production.

Deep learning, convolutional neural networks, transfer learning, image classification, Rice Leaf Disease, Classification, Detection, YOLO, LeafNet, Accuracy

1. A. Bhowmik, M. Sannigrahi, D. Chowdhury, and D. Das, ‘‘RiceCloud: A cloud integrated ensemble learning based Rice leaf diseases prediction system,’’ Proc. IEEE 19th India Council Int. Conf. (INDICON), Nov. 2022, pp. 1–6. [Google Scholar] [Crossref]

2. O. V. Putra, N. Trisnaningrum, N. S. Puspitasari, A. T. Wibowo, and E. Rachmawaty, ‘‘HiT-LIDIA: A framework for Rice leaf disease classification using ensemble and hierarchical transfer learning,’’ Lontar Komputer, vol. 13, no. 3, p. 196, Dec. 2022. [Google Scholar] [Crossref]

3. H. Zhang, ‘‘Attention-based feature enhancement for Rice leaf disease recognition,’’ Proc. 2nd Int. Conf. Artif. Intell., Autom., High-Perform. Comput. (AIAHPC), Nov. 2022, pp. 1234806-1–1234806-9. [Google Scholar] [Crossref]

4. N. Krishnamoorthy, L. V. N. Prasad, C. S. P. Kumar, B. Subedi, H. B. Abraha, and S. E. Sathishkumar, ‘‘Rice leaf diseases prediction using deep neural networks with transfer learning,’’ Environ. Res., vol. 198, Jul. 2021, Art. no. 111275. [Google Scholar] [Crossref]

5. S. Ghosal and K. Sarkar, ‘‘Rice leaf diseases classification using CNN with transfer learning,’’ IEEE Calcutta Conf. (CALCON), Kolkata, India, 2020, pp. 230–236. [Google Scholar] [Crossref]

6. R. R. Patil and S. Kumar, ‘‘Rice-fusion: A multimodality data fusion framework for Rice disease diagnosis,’’ IEEE Access, vol. 10, pp. 5207–5222, 2022. [Google Scholar] [Crossref]

7. L. Wei, Y. Luo, L. Xu, Q. Zhang, Q. Cai, and M. Shen, ‘‘Deep convolutional neural network for Rice density prescription map at ripening stage using UAV-based remotely sensed images,’’ Remote Sens., vol. 14, no. 1, p. 46, Dec. 2021. [Google Scholar] [Crossref]

8. J. Liu and X. Wang, ‘‘Plant diseases and pests detection based on deep learning: A review,’’ Plant Methods, vol. 17, no. 1, p. 22, Dec. 2021. [Google Scholar] [Crossref]

9. P. Barré, B. C. Stöver, K. F. Müller, and V. Steinhage, ‘‘LeafNet: A computer vision system for automatic plant species identification,’’ Ecological Informat., vol. 40, pp. 50–56, Jul. 2017. [Google Scholar] [Crossref]

• Advanced Techniques for Fake News Detection on Twitter Using NLP and AI: A Comprehensive Review
• Review of Self Compacting Geopolymer Concrete Using Slag Sand as Fine Aggregate
• Performance of Local Construction Contractors – Case Study of Registered Contractors in Monrovia, Liberia
• Cross-Cultural Perspectives on Innovation Management in Multinational Organizations
• Modeling of Reaction Between Dissolved Oxygen (DO) And Biological Oxygen Demand (BOD) in Degradation River