Publication

Abstract : In the field of augmented reality, deep learning techniques play a crucial role in enhancing the accuracy of crop disease detection. Cassava, a vital staple for millions in Sub-Saharan Africa and Southeast Asia, stands to gain significantly from the transformative potential of virtual and augmented reality in detecting leaf diseases. These immersive solutions provide innovative tools for farmers, allowing real-time disease detection and enabling timely management actions. Integrating augmented reality into cassava farming practices holds promise for healthier plants, improved yields and enhanced crop quality. Despite various methods proposed for cassava leaf disease detection, existing approaches have limitations, particularly in accuracy. Addressing these challenges, this research introduces a novel real-time solution as augmented reality and virtual reality-based cassava leaf disease detection using Modified Pyramidal Convolutional Shuffle binary Attention Residual network with Advanced Harris Hawk optimization algorithm (MPCSAR-AHH) model. The MPCSAR-AHH model demonstrates remarkable accuracy, with 99.02% precision, 97.55% recall and F1-score, outperforming previous methods. Additionally, the inclusion of a user-friendly graphical interface in augmented reality enhances the interpretability of disease detection results for farmers. Overall, the MPCSAR-AHH approach represents a significant advancement in cassava leaf disease detection and management, providing a robust and efficient solution