by
,
,
,
,
,
,
Digital Intelligence in Agriculture | Volume 1, Issue 2: 79-95, 2025 | DOI: 10.62762/DIA.2025.672386
Abstract
Maize productivity in India, a major global producer, is severely threatened by leaf diseases. Accurate identification of Common Rust (CR), Northern Corn Leaf Blight (NCLB), and Gray Leaf Spot (GLS) remains challenging with traditional methods. This study evaluated traditional and ensemble-based classifiers for classifying these diseases alongside healthy (HL) leaves. Using accuracy, precision, recall, and F1-score, we assessed k-NN, DT, RF, ETs, AdaBoost, SGD, GB, XGBoost, LightGBM, and a Stacking model on a four-class dataset. Ensemble methods demonstrated clear superiority. The Stacking model achieved the highest accuracy (98.50%), followed by LightGBM (98.46%) and XGBoost (98.01%). Among... More >
Graphical Abstract
