LeafDNet: Transforming Leaf Disease Diagnosis Through Deep Transfer Learning.

Area of Science:

• Agricultural Science
• Computer Science
• Plant Pathology

Background:

• Accurate plant disease detection is crucial for agricultural productivity.
• Traditional methods are labor-intensive and often inaccurate.
• Need for precise, automated solutions in modern agriculture.

Purpose of the Study:

• To develop an advanced deep transfer learning model for identifying plant diseases.
• To improve the accuracy and efficiency of disease detection in key crops.
• To provide a scalable solution for plant health management.

Main Methods:

• Utilized an enhanced Xception architecture with additional convolutional and dense layers.
• Incorporated advanced regularization and dropout techniques for optimization.
• Trained and validated the model on a dataset of 5491 plant leaf images across four disease categories.

Main Results:

• Achieved 98% accuracy, 99% precision, 98% recall, and 98% F1-score.
• Demonstrated superior performance compared to traditional and other deep learning methods.
• The enhanced model effectively captured subtle disease patterns.

Conclusions:

• The proposed deep learning framework offers a highly accurate and efficient solution for early plant disease detection.
• This technology supports sustainable agricultural practices and enhances plant health management.
• The model shows significant potential for real-world agricultural applications.