GLNet: global-local feature network for wheat leaf disease image classification.

Area of Science:

• Agricultural Science
• Computer Vision
• Machine Learning

Background:

• Traditional convolutional neural networks struggle with multi-scale feature perception and global information understanding for wheat leaf disease classification.
• Accurate identification of wheat diseases is crucial for crop management and food security.

Purpose of the Study:

• To address limitations in existing models for wheat leaf disease image classification.
• To propose a novel network architecture, GLNet, for enhanced feature perception and classification accuracy.

Main Methods:

• Developed a global-local feature network (GLNet) with a unique architecture.
• Processed global and local feature blocks in parallel.
• Integrated multi-scale features using a feature fusion block.

Main Results:

• GLNet achieved comprehensive multi-scale feature capture in wheat leaf images.
• The model demonstrated excellent performance and accuracy in classifying wheat leaf diseases in real-world scenarios.
• The innovative design significantly improved the model's understanding of disease image features.

Conclusions:

• GLNet offers a significant advancement in wheat leaf disease image classification.
• The proposed architecture provides new insights and effective tools for complex image classification tasks.
• This work contributes to the application of AI in precision agriculture.