Paper Contents
Abstract
The widespread occurrence of plant diseases poses a serious risk to global food security and agricultural sustainability. This study introduces an advanced system designed for the early detection, severity assessment, and customized treatment of plant diseases. Utilizing sophisticated deep learning models such as ResNet-50, DenseNet-201, VGG16, VGG19 and EfficientNetV2, the system accurately identifies diseases from leaf images, achieving a test accuracy of up to 97.02%. Disease severity is evaluated using an area-based method that categorizes the condition into stages - mild, moderate, severe, and very severe based on the proportion of infected leaf area. The system also incorporates treatment recommendation algorithms that provide targeted solutions based on disease type and severity, encouraging sustainable and efficient farming practices. The framework is trained on the New Plant Diseases dataset, which comprises data for different crops, including tomatoes, apples, and corn and 14 other plants, covering 20 different diseases. Real-time detection capabilities are enabled through an easy to-use platform. The experimental results show the robustness of the system in different environmental conditions, providing a scalable and cost-effective means of reducing crop losses and advancing precision agriculture to support data-driven decision making and more sustainable farming practices.
Copyright
Copyright © 2025 Harshitha GN. This is an open access article distributed under the Creative Commons Attribution License.
