Classification of tomato leaf disease using Transductive Long Short-Term Memory with an attention mechanism

Area of Science:

• Agricultural Science
• Computer Science
• Machine Learning

Background:

• Tomato cultivation faces challenges due to leaf diseases, impacting yield and quality.
• Accurate and early disease detection is crucial for effective crop management and preventing losses.
• Existing deep learning models often struggle with classification complexity and limited computational power.

Purpose of the Study:

• To develop an advanced deep learning model for precise tomato leaf disease classification.
• To overcome the limitations of single-architecture models in handling complex image data.
• To enhance the efficiency and accuracy of automated disease detection systems in agriculture.

Main Methods:

• Utilized a novel Transductive Long Short-Term Memory (T-LSTM) model integrated with an attention mechanism.
• Employed transductive learning principles to leverage dataset characteristics for accurate predictions.
• Incorporated U-Net for image segmentation and VGG-16 for feature extraction, followed by T-LSTM classification.
• Pre-processed data from the PlantVillage dataset using image resizing, color enhancement, and data augmentation.

Main Results:

• The proposed T-LSTM classifier achieved a high accuracy of 99.98% in tomato leaf disease classification.
• Demonstrated superior performance compared to existing convolutional neural network (CNN) models with transfer learning and IBSA-NET.
• The attention mechanism effectively focused on relevant image sequence parts for improved classification.

Conclusions:

• The T-LSTM model with an attention mechanism offers a highly accurate and efficient solution for tomato leaf disease classification.
• This approach significantly improves upon current deep learning methods, paving the way for better agricultural disease management.
• The findings highlight the potential of transductive learning and attention mechanisms in complex image classification tasks within agriculture.