A Novel Detection on Wheat Disease through CL and RGB Filters by LDA and QDA

Wheat is essential everywhere. Wheat leaf diseases hinder growth. Wheat leaf disease identification is crucial to wheat quality and agriculture. This work presents an integrated machine learning strategy to improve wheat leaf disease identification: Colour Layout Filter with Linear Discriminant Analysis, Colour Layout Filter with Quadratic Discriminant Analysis, RGB Filter with Linear Discriminant Analysis and RGB Filter with Quadratic Discriminant Analysis can identify damaged wheat leaves. The agricultural autonomous leaf infection detection system uses images, processing, feature extraction, selection, and learning. This technology helps farmers quickly and reliably diagnose plant illnesses. Automatic leaf disease detection speeds crop diagnosis. This study’s Linear Discriminant Analysis Color Layout Filter classifies wheat diseases well. LDA-CLF is most accurate at 88.33%. QDA-RGBF has 80% accuracy. CLF with LDA 0.88 is ideal. RGBF QDA accuracy is 0.80, poor. LDA-CLF recalls 0.88. RGBF with QDA recall is 0. S0, low. With 0.66 kappa, CLF and LDA lead. RGBF with QDA has lowest kappa (0.49). Our best model is CLF’s0.88 LDA F-Measure. QDA-enhanced RGBF has 0.80 F-Measure. LDA-CLF has the highest MCC at 0.66. RGBF QDA MCC lowest is 0.5. 0.93 CLF-LDA ROC. RGBF’s LDA-based ROC is 0. S4. RGBF QDA models have the highest PRC (0.92). RGBF-LDA has the lowest PRC (0.78). Linear Discriminant Analysis-based Color Layout Filters outperformed other models in this study.