Leveraging Pre-trained CNNs for Efficient Feature Extraction in Rice Leaf Disease Classification

TL;DR

The study targets accurate rice leaf disease classification by evaluating the impact of integrating hand-crafted feature extraction with pre-trained CNNs. It demonstrates that Histogram of Oriented Gradients (HOG) significantly enhances performance, especially with EfficientNet-B7, achieving up to 97% accuracy, while Local Binary Patterns (LBP) provide limited gains. Grad-CAM is employed to reveal disease-specific regions, increasing interpretability of the models. Overall, the findings suggest that combining HOG-based features with powerful pre-trained CNNs can substantially improve image-based agricultural disease identification and reliability for practical deployment.

Abstract

Rice disease classification is a critical task in agricultural research, and in this study, we rigorously evaluate the impact of integrating feature extraction methodologies within pre-trained convolutional neural networks (CNNs). Initial investigations into baseline models, devoid of feature extraction, revealed commendable performance with ResNet-50 and ResNet-101 achieving accuracies of 91% and 92%, respectively. Subsequent integration of Histogram of Oriented Gradients (HOG) yielded substantial improvements across architectures, notably propelling the accuracy of EfficientNet-B7 from 92\% to an impressive 97%. Conversely, the application of Local Binary Patterns (LBP) demonstrated more conservative performance enhancements. Moreover, employing Gradient-weighted Class Activation Mapping (Grad-CAM) unveiled that HOG integration resulted in heightened attention to disease-specific features, corroborating the performance enhancements observed. Visual representations further validated HOG's notable influence, showcasing a discernible surge in accuracy across epochs due to focused attention on disease-affected regions. These results underscore the pivotal role of feature extraction, particularly HOG, in refining representations and bolstering classification accuracy. The study's significant highlight was the achievement of 97% accuracy with EfficientNet-B7 employing HOG and Grad-CAM, a noteworthy advancement in optimizing pre-trained CNN-based rice disease identification systems. The findings advocate for the strategic integration of advanced feature extraction techniques with cutting-edge pre-trained CNN architectures, presenting a promising avenue for substantially augmenting the precision and effectiveness of image-based disease classification systems in agricultural contexts.

Figures (6)

• Figure 1: Proposed Workflow for Rice Disease Classification
• Figure 2: Visual representation of rice diseases. (Left to right): Brown spot, Leaf blast, and Neck blast
• Figure 3: Visualization of HOG Features for Brown Spot Disease in Rice Leaves.
• Figure 4: Visualization of LBP Features: Leaf Blast Disease in Rice Leaves.
• Figure 5: Comparison of Classification Accuracies among Various Models with Different Feature Extraction Techniques.