Deep Learning Model for Classifying and Evaluating Soybean Leaf Disease Damage

Area of Science:

• Agricultural Science
• Computer Science
• Plant Pathology

Background:

• Soybean crops are susceptible to various damaging factors, impacting yield.
• Accurate damage assessment is crucial for effective crop management and yield forecasting.
• Existing deep learning models (DLMs) are limited to binary healthy/unhealthy classification.

Purpose of the Study:

• To develop a novel DLM for predicting and categorizing soybean leaf damage severity into five levels.
• To provide a comprehensive solution for distinguishing between healthy and unhealthy soybean leaves.
• To support tailored pesticide application and enhance yield projections.

Main Methods:

• Trained a novel DLM on 2930 near-field soybean leaf images.
• The model quantifies damage severity across multiple levels.
• Evaluated performance using accuracy, precision, recall, and F1-score.

Main Results:

• The DLM accurately predicts and classifies soybean leaf damage severity.
• The model effectively distinguishes between healthy and unhealthy leaves.
• Achieved high performance metrics, indicating robust damage assessment capabilities.

Conclusions:

• This research presents a robust DLM for soybean damage assessment.
• The model supports informed agricultural decisions based on specific damage levels.
• Enhances crop management strategies and boosts overall agricultural productivity.