Plant Leaf Disease Detection and Classification using Image Processing Techniques

Agriculture: the backbone of livelihood in India, where a significant portion of the economy is dependent on agriculture. And with a burgeoning population, agricultural systems come under pressure to supply enough, high-quality yields to guarantee food security and economic stability. Diseases of plants that impose highly visible constraints on growth are one of the principal villains causing loss of productivity in agriculture, resulting in major deficits, both in farm productivity and farmer income. Hence, a timely and accurate identification of these diseases is crucial. Therefore, this work offers an approach based on image processing for identifying and classifying the diseases on the leaf of a plant coupled with machine learning algorithms. To lay a foundation of proven techniques, a thorough review of current research was performed. The technique involves deploying seven machine learning classifiers, which are Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Classification and Regression Trees (CART), Random Forest (RF), Gaussian Naïve Bayes (GNB), Logistic Regression (LR), and Linear Discriminant Analysis (LDA). Performance was measured in terms of precision, recall, F1-score, specificity, and accuracy. Random Forest model performed best of all the classifiers with 98.12% accuracy level; this highlights that ensemble methods in general pull ahead of others in real-world disease detection applications. The findings highlight the power of AI-based tools in facilitating effective, scalable, and sustainable agricultural practices.