Wheat Disease Detection Using Transfer Learning Techniques

Wheat stands as a crucial staple crop for a substantial portion of the global population, contributing significantly to food security.However, the productivity and expansion of wheat cultivation face substantial challenges due to the prevalence of diseases, resulting in considerable annual crop losses.Nowadays, deep learning methods have become major in the identification of leaf diseases.The study proposes the techniques that mainly concentrating on transfer learning (TL) architectures, to advance agricultural research.Various TL architectures, such as VGG16, ResNet50, Squeeze Net, and VGG19, are explored for disease detection in wheat plants.The methodology involves preprocessing of leaf images, utilizing TL architectures to extract the features of the leaf.Subsequently, TL architectures are fine-tuned using these segmented images, and the fully connected layers of the combined architecture of VGG19 and RESNET50 are employed for disease classification.The model focuses on all diseases caused by fungi and bacteria in wheat plants.The analysis confirms that the developed model outperforms existing counterparts, highlighting its efficacy in advancing wheat leaf disease detection.This project contributes to empowering farmers with innovative tools for accurate and early disease detection, ultimately safeguarding wheat crop yield and quality.

Перевод пока недоступен