Plant Disease Detection and Classification Through Deep Learning and Transfer Learning

The innovative progress of a nation, particularly in India, largely depends on its agricultural sector, which plays a crucial role in ensuring food security for the population. However, plant diseases pose a significant threat to agricultural productivity, making their early and accurate detection a critical concern. Traditional methods for plant disease identification rely on manual inspection, which is time-consuming, subjective, costly, and requires extensive expertise and information about various plant diseases. This highlights the urgent need in India for automating conventional techniques through the adoption of advanced deep learning (DL) technologies.In this study, we propose an automated approach for plant disease detection in India using DL models and transfer learning techniques applied to plant images. A custom CNN model was implemented for image classification, along with transfer learning models such as ResNet50, EfficientNetB3, VGG19, and MobileNetV2. The models were evaluated using performance metrics such as accuracy and loss. Among the tested models, ResNet50 achieved the highest accuracy of 97.7%, demonstrating its effectiveness for addressing agricultural challenges in the Indian context.