Multi-model machine learning for automated identification of rice diseases using leaf image data.

Area of Science:

• Agricultural Science
• Computer Science
• Plant Pathology

Background:

• Rice is a vital global food source, threatened by diseases causing significant yield loss.
• Current disease identification methods are slow, labor-intensive, and require expert knowledge.
• Automated, rapid, and cost-effective disease detection is crucial for modern agriculture and food security.

Purpose of the Study:

• To develop a novel hybrid deep learning and machine learning framework for automated rice plant disease identification.
• To improve the efficiency and accuracy of rice disease detection from leaf images.
• To reduce farmer reliance on manual inspection and support timely agricultural interventions.

Main Methods:

• Extracted deep features from rice leaf images using pre-trained Convolutional Neural Network (CNN) models (MobileNetV2, DarkNet19, ResNet18).
• Classified extracted features using Support Vector Machine (SVM) classifiers with various kernel functions.
• Employed a 10-fold cross-validation technique to ensure model reliability and performance evaluation.

Main Results:

• Achieved a high classification accuracy of 98.61% using an SVM classifier with a medium Gaussian kernel.
• Demonstrated excellent performance with a specificity of 98.85% and sensitivity of 97.25%.
• The framework proved computationally efficient and scalable for handling larger datasets.

Conclusions:

• The proposed hybrid framework offers a dependable and efficient solution for accurate rice leaf disease identification.
• Automated detection significantly reduces the need for manual inspection, empowering farmers with timely intervention capabilities.
• This technology supports sustainable rice production and contributes to global food security.