Early Detection and Classification of Tomato Leaf Disease Using High-Performance Deep Neural Network.

Area of Science:

• Agricultural Science
• Computer Vision
• Plant Pathology

Background:

• Tomato crop yield is significantly impacted by leaf diseases, necessitating accurate and timely diagnosis.
• Early identification of tomato plant diseases is vital for mitigating crop loss and improving yield quality.

Purpose of the Study:

• To develop and evaluate a Convolutional Neural Network (CNN) model for the accurate classification of tomato leaf diseases.
• To provide farmers with a tool for early-stage disease detection to enhance crop management.

Main Methods:

• A dataset of 3000 tomato leaf images, encompassing nine distinct diseases and healthy samples, was utilized.
• Image preprocessing and segmentation were performed, followed by CNN model training with hyper-parameter tuning.
• The CNN model extracted features such as color, texture, and edges for disease classification.

Main Results:

• The proposed CNN model demonstrated a high prediction accuracy of 98.49% in classifying tomato leaf diseases.
• The model effectively identified various disease indicators from image characteristics.

Conclusions:

• Convolutional Neural Networks offer a powerful approach for automated tomato leaf disease diagnosis.
• The developed model shows significant potential for practical application in agriculture, aiding farmers in disease management.