PENERAPAN METODE SUPPORT VECTOR MACHINE (SVM) PADA KLASIFIKASI JENIS CENGKEH BERDASARKAN FITUR TEKSTUR DAUN

Abstract

Leaves are a very important plant component because they play an important role in differentiating plant species, including clove plants. Currently, the identification of clove species, namely Afo, Siputih, and Zanzibar, relies on manual observation of the characteristics of the fruit and flowers, which can take a long time, especially considering the long fruiting period of the clove plant. To answer this problem, the authors conducted a study to classify the three types of clove leaves based on the characteristics and texture of the Gray gray-level co-occurrence Matrix (GLCM), which includes four parameters: Contrast, Correlation, Energy, and Homogeneity. 
 The Support Vector Machine (SVM) classification algorithm processes extracted feature values and accurately class leaves. This study achieves the highest accuracy of 56.67% on an image size of 250x250 pixels and 48.33% on an image size of 150x150 pixels using 150 training data and 60 test data. These results indicate the potential of automatic leaf classification in efficiently identifying clove plant species.
 Keywords : Clove, Leaf, Processing, Texture, SVM