Optimizing Paddy Leaf Disease Detection: A Comparative Analysis of Advanced Deep Learning Approaches

An adequate level of agricultural productivity is generally necessary for human life.The paddy plant is regarded as a key planting crop that can raise the economic prosperity of the nation.Numerous diseases currently threaten to reduce the productivity of paddy crops.The major causes of this type of disease on paddy crops are bacteria, fungi, viruses, and some dangerous insects.Crop diseases that impact rice crops in their early stages might have an impact on the entire crop cultivation process.Farmers carried out the manual detection of diseases in the early days of agriculture.Manual disease detection and classification require a significant amount of effort and often result in specific inaccuracies.Image processing and artificial intelligence are developing methods for identifying and categorizing various diseases to address the challenges faced in manual disease identification.Recent advancements in deep learning have enabled more accurate and efficient detection methods.This study introduces a hybrid deep learning approach for paddy leaf disease detection, integrating advanced neural network architectures to enhance classification performance.The proposed models include InceptionV3+GRU, EfficientNet B0+Capsule Network, and DenseNet 121 with a Transformer model.These architectures were selected based on their capability to extract complex spatial and contextual features, improving disease classification accuracy.These combinations address the limitations of conventional CNNs.The comparative analysis is conducted on a curated dataset containing five distinct paddy leaf diseases.The experimental results demonstrate that DenseNet 121 with a Transformer model outperforms other models, achieving an accuracy of 98.95%.The integration of DenseNet 121 with Transformer leverages its ability to capture spatial dependencies and long-range contextual relationships, leading to superior classification results.This study highlights the effectiveness of combining convolutional neural networks with advanced architectures for precise paddy leaf disease detection, offering valuable insights for precision farming and disease management strategies.