A novel GCL hybrid classification model for paddy diseases

Area of Science:

• Agricultural Science
• Computer Science
• Artificial Intelligence

Background:

• Growing global population necessitates increased agricultural output.
• Advancements in AI, particularly computer vision and deep learning, offer solutions for industry challenges.
• Accurate and reliable disease detection is crucial for crop yield and food security.

Purpose of the Study:

• To introduce a novel hybrid neural network model, GCL, for classifying paddy diseases.
• To enhance the accuracy and reliability of AI-based disease classification systems.
• To leverage data augmentation techniques for improved model performance.

Main Methods:

• Developed a hybrid model (GCL) fusing Generative Adversarial Network (GAN), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM).
• Utilized GAN for dataset augmentation to improve training data diversity.
• Employed CNN for feature extraction and LSTM for disease classification from paddy leaf images.

Main Results:

• The GCL model demonstrated suitability for paddy disease classification.
• Achieved a testing accuracy of 97% in classifying bacterial blight, leaf smut, and rice blast.
• The hybrid approach effectively combined data augmentation with deep learning for disease identification.

Conclusions:

• The GCL model offers a highly accurate and reliable solution for paddy disease classification.
• The integration of GAN, CNN, and LSTM provides a robust framework for agricultural disease detection.
• The GCL model has the potential for expansion to classify a wider range of crop diseases.