Identification of Herbal Plants Using Morphology Method and K-Nearest Neighbour Algorithm (KNN)

Abstract

The large number of herbal plants and the lack of public knowledge about the types of herbal medicines make it difficult for people to distinguish the types of herbal medicinal plants. So, to provide information for public, an identification system that could identify and recognize a herbal medicinal plant is needed. Leaves identificiation is based on the Morphological method with K-Nearest Neighbor algorithm. Leafs data that will be used in this study is Basil Leaves, Orthosiphon leaves, Painted Nettle leaves, Betel leaves, and Teak leaves which are categorized as flowering plants from the Lamiaceae or Labiatae family for the classification process. Image processing techniques will be used to obtain the pattern of the shape and characteristics of each leaf which is known as the recognition of leaf morphological features. This study using 200 leaves data and separated into 2 types, with 70% basic data and 30% test data. With Morphological Feature Extraction using area and parameters as well the compactness and metric features of the leaves. The accuracy of the classification results using the confusion matrix testing method with K3 values obtained by 80%, K5 and K7 values obtained accuracy of 76.67% and K9 values showing accuracy of 78, 67%.