Wheat leaf diseases classification and severity analysis using HT-CNN and Hex-D-VCC-based boundary tracing mechanism

Area of Science:

• Agricultural Science
• Computer Vision
• Plant Pathology

Background:

• Wheat is a vital global crop, susceptible to significant fungal diseases.
• Early and accurate disease identification is crucial for enhancing wheat yield and minimizing losses.

Purpose of the Study:

• To propose a novel HT-CNN model for predicting wheat leaf disease and its severity.
• To improve the accuracy and efficacy of wheat disease detection.

Main Methods:

• Image preprocessing: RGB to HSV conversion, V-component filtering and enhancement.
• Segmentation using weighted Canberra distance-based K-means (WCD-K means) and binarization.
• Feature extraction with Hex-D-VCC, optimal feature selection using baker's map-based Harris hawks optimization (BM-HHO).
• Classification and severity calculation using the proposed HT-CNN model.

Main Results:

• The proposed HT-CNN model demonstrated higher accuracy and efficacy compared to other methods.
• The integrated approach effectively identifies diseased regions and classifies disease severity.

Conclusions:

• The developed HT-CNN model offers a robust solution for early wheat leaf disease detection and severity assessment.
• This approach has the potential to significantly contribute to sustainable agriculture and food security.