Investigation on Leaf Disease Diagnosis in Rice Plant using Machine Learning Approaches

The discovery of crop diseases in the early phase is vital in the agricultural field. This helps in treating the crop with necessary actions to avoid the disease’s spread in the early stages. Research shows crop yields and quality may generally be improved by utilizing machine learning techniques. This work examines the performance of various machine learning models that helps to identify an efficient model to diagnose crop diseases in the early phase thus reducing the time and cost expense. Initially, the input images are collected from the Rice Leaf Disease Image dataset and pre-processed for further processing. The feature extraction process makes use of the pre-processed image and extracts useful insights from it. These extracted features are then given into the machine learning models which predict the target value. The various machine learning algorithms used in this research work include K-Nearest Neighbors (K-NN), Support Vector Machine (SVM), Naive Bayes (NB), Decision Tree (DT), and Random Forest (RF). The predicted results are compared against the actual values for every model which provides the performance metrics of these models. According to the computed performances, the Random Forest Classifier provides the highest accuracy in classifying whether the rice plant leaf has the disease or not.