Oryza Sativa Disease Diagnosis Through Particle Swarm Optimized Shifted Windows Transformer

One of the primary problems farmers encounters is the frequent occurrence of diseases that specifically target the leaves of rice plants. This makes it extremely difficult for farmers to supply the rising demand for food needed to support an expanding population. The agricultural sector suffers substantial losses due to rice diseases, which greatly affect food production. Traditional methods of identifying diseases in rice leave have limitations in accuracy and computational efficiency. To tackle these challenges, this study introduces a novel approach: a rice leaf disease detection system employing the Shifted Windows Transformer (SWIN). This system represents a significant advancement in rice disease classification. Unlike conventional methods such as Convolutional Neural Networks (CNNs), which may struggle to capture the full context efficiently, Shifted Windows leverages hierarchical transformer blocks to better understand the complexities of images. By utilizing this architecture, the proposed system offers improved precision and computational efficiency in identifying rice leaf diseases from captured images from rice leaves giving high accuracy.