,
Research Scholar,Department of Computer Science, Karpagam Academy of Higher Education , Coimbatore, Tamil Nadu , India
Professor and Head, Department of Computer Science, Karpagam Academy of Higher Education , Coimbatore, Tamil Nadu , India
This research introduces a sophisticated deep learning framework for automatically identifying diseases in rice leaves by combining the Inception V3 architecture with spatial attention mechanisms. Rice is one of the most important foods in the world, in terms of food security and agricultural economics, so the establishment of efficient disease surveillance systems is now necessary to support sustainable farming activities. This is particularly important in achieving SDG 2 (Zero Hunger), which aims to ensure access to sufficient food, and SDG 12 (Responsible Consumption and Production), which promotes sustainable farming practices. Convolutional neural networks have been successfully used to classify plant diseases in the past, but traditional convolutional neural network models are sometimes not capable of ranking the most diagnostically relevant features. The attention mechanisms used in this methodology defeat this challenge by integrating the Inception V3 framework. In particular, the spatial attention aspect guides the model to areas that are of disease-specific features. The study used a dataset of 18,160 images of the rice leaf, which included nine separate disease types and controls that were selected through the Kaggle and Plant Village datasets. Results showed that the attention-enhanced hybrid model reached 92.98% accuracy in classification tasks, surpassing the standard Inception V3 baseline in fewer than 12 training epochs. Significant improvements were also noticed in the cases of distinguishing diseases having similar visual appearance, especially in early and late blight conditions. The model showed strong performance across all ten rice disease categories that were tested, reaching its best validation accuracy of 92.98% during the 12th training epoch. These findings indicate that there is indeed a benefit of incorporating attention processes into the InceptionV3 architecture for the task of disease detection in agricultural crops.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
0
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.