Enhancing sustainable agriculture using attention convolutional bidirectional Gated recurrent based modified leaf in wind algorithm: Integrating AI and IoT for efficient farming

Sustainable agriculture is essential for ensuring global food security while mitigating environmental impacts. The possibilities of using remote sensing data and artificial intelligence in agricultural practices emphasize optimizing resource use, minimizing waste, and fostering resilient farming systems to adapt to changing climate conditions in the agriculture field. Multiple studies employed in utilizing remote sensing data and AI for diagnosing disease and environmental monitoring but they face challenges due to factors such as distortions and changes in climates like huge rainfall and extreme droughts affecting the farming environment. Therefore this article develops a novel Attention Convolutional Bidirectional Gated Recurrent based Modified Leaf in Wind Algorithm for assessing the disease of the plants and environmental monitoring. The algorithm leverages diverse datasets including PlantVillage, plantDoc, Soil Type, Advanced IoT Agriculture, and IDADP, and robust data preprocessing techniques such as normalization, standardization, and imbalanced data handling are essential for refining dataset integrity and optimizing model performance. Additionally, the developed model incorporates a convolutional neural network for spatial feature extraction, bidirectional gated-recurrent units for sequential context modeling, and attention mechanisms fuse the Convolutional Neural Network and bidirectional gated-recurrent units, focused on increasing the activity of the proposed network to obtain optimal results, by applying weighting model to each time steps. Moreover, to improve feature integration and optimize model performance, the proposed algorithm incorporates Modified Leaf in Wind optimization strategies. Through experimental validation, the proposed method procures the best performance in four scenarios with a precision of 97.5 % for SC1, 98.5 % for SC2, 96.9 % for SC3 %, and 97.6 % for SC4. The proposed model empowers farmers to make data-driven decisions that enhance productivity.