Image Classification of Rice Leaf Diseases Using Random Forest Algorithm
Panuwat Mekha, Nutnicha Teeyasuksaet
Abstract
The problem of rice diseases around the world make to damage and fall into a large number of rice. Caused by many of types, such as; fungi, Bakteri and Viruses. which are the main causes of rice disease affected to farmers. The classification of rice can be classified into several methods. In this research, image classification is used to classify the data set of rice leaf diseases, such as; Brown Spot Rice disease (BSR), Brown Spot Rice disease (BSR), Bacterial Leaf Blight disease (BLB), which is the rice leaf disease with severe outbreaks around Thailand. Moreover, image processing technology in the classification types of rice leaf disease, such as; Random forest classification algorithm, Decision tree classification algorithm, Gradient Boosting classification algorithm and Naïve-Baye classification algorithm, which is measured by the accuracy, precision and recall of each algorithms. The best result of performance in the image classification of rice leaf diseases is random forest algorithm equal to 69.44 percent.