Crop Disease Analysis with Different Classification Models using Orange Data Mining Tool

Agriculture is one of the principal economic activities of the India on agriculture for their livelihood. This study presents a crop disease prediction system utilizing the Orange data mining tool to analyze image datasets of diseased plant leaves from various fruits and vegetables, including apple, strawberry, blueberry, cherry, corn, and potato. The system employs machine learning algorithms to classify and predict crop diseases through the analysis of leaf images. Includes an image data set comprising various healthy and unhealthy crop leaves. Image dataset containing different healthy and unhealthy crop leaves was collected and preprocessed. Multiple classification models, including Image Embedding, Logistic Regression, Random Forest, KNN, Tree, SVM,Neural Networks [1] Scatter Plot, ROC Analysis, Confusion Matrix,are implemented in the Orange environment. The model underwent training using 80% of the dataset, while the remaining 20% was utilized for validation. The results indicated that the Supervised Learning model attained the highest level of accuracy in recognizing and categorizing crop diseases based on leaf images. This system offers farmers an effective means for the early detection of diseases, facilitating prompt intervention and potentially minimizing crop losses. Future efforts will concentrate on broadening the image dataset and integrating more crop varieties to improve the system's adaptability and precision.