Deep Identification of Plant Diseases

Plant diseases pose significant challenges to global crop production, impacting the economy. Innovative agricultural solutions that integrate the Internet of Things and machine learning have emerged to address this issue for early discovery of plant pathogens. While convolutional neural networks (CNNs) have been widely used for plant disease detection, recent advancements in deep learning have introduced vision transformers (ViTs) as highly effective models for classification tasks in various vision-based applications. Researchers have started exploring the potential of ViTs for plant pathology applications. This paper proposes a hybrid model combining the strengths of Pyramid ViT (PVT) and the powerful feature extraction capability of VGG-16 for disease identification. The model demonstrates its efficiency in identifying numerous plant diseases across different crops. By comparing eight modern techniques on PlantVillage dataset, the proposed model outperforms all others, achieving an impressive accuracy of 98.51% and a precision of 97%.