Early disease detection in rubber plants (Hevea brasiliensis) is challenging, requiring expert knowledge and experience to confirm diseases, which is time-consuming and costly. Therefore, this study aims to develop a disease detection and prediction system using image processing techniques and artificial intelligence methods. Four types of diseases were identified: three leaf diseases (Oidium powdery mildew, Corynespora, and Collectotrichum) and one root disease (white root disease) with three stages (light, moderate, and severe). Samples were collected from rubber plantations in Tabalong, South Kalimantan, totaling 450 images. The dataset was modeled based on expert labeling. GLCM was used for texture extraction, with six selected features: contrast, correlation, energy, homogeneity, entropy, and inverse difference moment. The utilization of ANFIS and RBFNN provides a powerful and flexible approach to plant disease detection. These methods learn from training data and adjust their par.....

199 hits