<![CDATA[Identification of diseases from images of plants is one of the interesting research areas in the agriculture field, for which machine learning concepts from the computer science field can be applied. This article presents a prototype system for the detection and classification of rice diseases based on images of infected rice plants. This prototype system was developed after detailed experimental analysis of various techniques used in image processing operations. We consider three rice plant diseases: Bacterial Leaf Blight, Blast, and Tungro. We used the Otsu method to remove the background. To enable accurate extraction of features, we combined Gabor and Sobel techniques. In the classification process, we used five machine learning techniques: Random Forest (RF), Support Vector Machine (SVM), Nave Bayes (NB), and Quadratic Discriminant Analysis (QDA). We empirically evaluated these methods, achieving 77%, 50%, 60%, and 37% accuracy, respectively.]]>
Copyrights © 2024
