Melly Br Bangun, Fuzy Yustika Manik, Yeni Herdiyeni, Elis Nina Herliyana
Early identification of the disease is very important for the protection of plants. Image analysis can be applied to detect the disease in the leaves by utilizing the characteristic morphology in the identification process. Preprocess performed as the first step to change the image of the original leaf image into binary, further feature extraction based on morphological traits. The next perform feature extraction based on morphological characteristics such as convexity, solidity, elongation, roundness, rectangularity, eccentricity, and compactness. Each feature is used to measure the convexity, soliditas, extension, roundness, rectangular, elliptical and measure the compact of forms. Mathematical equations on a series of image pixels can be used to improve the aspect of shape and structure. So, it can be more easily recognized by the training and testing process used to build the moled classification using a support vector machine. The data used in this research is the image leaves jabon on seedbed phase affected by the disease spotting of leaves and leaf blight. The results obtained, the features of the morphology are able to form characteristic presentations of the symptoms of the symptom can quantitatively be explained by the shape and features of the SVM. As a method of classification it is able to identify the two disease leaves on jabon seeds with accuracy 87.5 %. © 2024 American Institute of Physics Inc.. All rights reserved.
Faculty of Science Education, Universitas Negeri Medan, Medan, Indonesia; Faculty of Computer Sciences and Information Technology, Universitas Sumatera Utara, Medan, Indonesia; Department of Computer Science, Bogor Agricultural University, Bogor, Indonesia; Department of Silvikultur, Bogor Agricultural University, Bogor, Indonesia
