Detection of Anthracnose and Black Sooty Mold Diseases in Mango Leaves Using Feature Extraction and Machine Learning Techniques

Agriculture has been the backbone of our country for ages. Many factors contribute to it, but the most common is maladies. The work focuses on Anthracnose and Black Sooty Mold diseases, which affect mango leaves and hinder production. This work comprises a machine learning-based method for diagnosing diseases in mango leaves. The segmentation of images is carried out using k-means segmentation, extracting features such as GLCM, HOG, SIFT, and Statistical Feature Matrix. Decision Tree, Support Vector Machine, Naive Bayes, K nearest Neighbors, and Random Forest are algorithms for disease classification in mango leaf. Experimental results demonstrated that Random Forest with Statistical Feature Matrix outperformed other algorithms without image segmentation. The proposed method shows promising results for promptly identifying diseases in mango leaves and strategies for disease management in agriculture.