Hops Plants Disease Detection using Feature Selection based BPSO-SVM

Hop plant is a plant used as a raw material for brewing beer, commonly grown in temperate climates. Some diseases, such as Downy Mildew and Powdery Mildew, can interfere with the cultivation process. The efforts that farmers have made have not gotten maximum results in this problem. Therefore, a more accurate and efficient detector is needed to reduce this problem. With the help of computer vision to identify diseases and machine learning to make correct decisions, we propose a disease detection system in hops. In this experiment, we compare the value of classification accuracy using feature selection and without feature selection. Based on these 2 methods, the Binary Particle Swarm Optimization (BPSO) feature selection and the Support Vector Machine (SVM) classifier resulted in a higher accuracy value of 95%, followed by the SVM classifier results without feature selection of 93%.