Rice Leaf Disease Detection using Fusion Optimization

Rice leaf disease detection is critical for maintaining crop yields and ensuring food security, especially in regions heavily reliant on rice cultivation. Traditional manual inspection methods are time-consuming, labor-intensive, and prone to error. This study presents an advanced deep learning-based ensemble model for accurate and efficient detection of three common rice leaf diseases: hispa, brown spot, and leaf blast. The proposed approach integrates feature fusion and ensemble learning techniques to leverage the complementary strengths of ResNet and DenseNet architectures. Robust data preprocessing and augmentation strategies are employed to enhance model performance. The ensemble model achieves superior accuracy compared to individual models, demonstrating the potential of ensemble learning for rice leaf disease detection. Experimental results on a field-captured dataset of 5,678images validate the model’s effectiveness, with an accuracy of $82 \%$ in disease classification. This research contributes to the development of automated disease detection systems, supporting sustainable agricultural practices and mitigating crop losses.