Classification of Apples based Leaf Using K-Nearest Neighbors and Moment Invariant Extraction

Abstract

Apples are one of several fruits that are widely favored and cultivated in various regions, in Indonesia many regions cultivate them and serve as the main source of livelihood. From the many types of apples that thrive, it is not difficult for the owner to recognize one type of apple that they planting, because everything looks the same. In previous studies a classification of apple plants using themethod K-Nearest Neighbor (K-NN) with RGB extraction on a dataset of 50 data and in this study an accuracy rate of 93.33% in terms of homogeneity features, texture features was 73, 33% and reaches 100% in terms of Red, Green, Blue (RGB) features. In another study, researchers used the method K-Nearest Neighbor (K-NN) with extraction using histrogram, producing 90% accuracy with 90 data training and 10 data testing using apples as research objects. In contrast to some previous studies, in this study will be carried out on the object image of an apple leaf with a dataset of 750 images. In this study, the author will examine the method K-Nearest Neighbor for the classification of apples based onleaf with the extraction feature moment invariant. The results of this study showed an accuracy of 74.93% for the classification of apple based on leaf which was carried out using the K-NN method with moment invariant for feature extraction method