BPSO based Feature Selection for Rice Plant Leaf Disease Detection with Random Forest Classifier

摘要

Recently, Machine Learning and computer vision have generated interest and have found new applications in engineering. In agriculture, “smart” systems have become important tools for detecting anomalies that decrease the quality and quantity in the harvest of agricultural products. This paper intended to detect three rice diseases, namely Brownspot, Bacterial Leaf blight, and Leaf smut, using the Random Forest Classifier technique of machine learning with image processing. The color moments are extracted for color features, while the Gabor Wavelet and Harris Corner methods are used for texture features extraction of PlantVillage Dataset images for rice plant leaf disease detection. The binary particle swarm optimization (BPSO) is then applied for the feature selection from the extracted features. Finally, Random Forest Classifier is used for the classification of extracted features to obtain the simulation results in terms of precision, sensitivity, and accuracy using a confusion matrix plot.