A growing population poses a significant challenge globally, especially for India, where increasing food demand requires an increase in agricultural production. Plant yields are negatively affected by pests, climate change, and traditional farming practices. A sustainable agricultural system relies on the control of biological agents, including bacteria, fungi, and viruses. The diagnosis of diseases is traditionally done by visual inspection, which can take a lot of time and requires expert judgment. It is imperative that recognition-based classifier models be utilized in order to overcome these limitations. Agricultural productivity has been reduced and crop quality has improved due pertaining to the most current developments in artificial intelligence (AI) and computer vision (CV). By reviewing the literature, we identify limitations in existing methods and propose an automated method for extraction of features from leaf images. Plant Disease Network (PDNet) uses XGBoost (Extreme Gradient Boosting) in order to identify and categorize leaf disease. The efficiency of state-of-the-art classifiers in the study, including Extra Trees, XGBoost, and PDNet, are compared using Ensemble Learning. In order to evaluate the results of evaluations, metrics such as F1 score, confusion matrix, recall, accuracy, and precision are used. In the results, the Support Vector Machine (SVM) combined with XGBoost is capable of producing 93.22% accuracy, which is superior to any other model tested.