Klasifikasi Jenis Anggur Berdasarkan Bentuk Daun Menggunakan Convolutional Neural Network Dan K-Nearest Neighbor

Abstract

. Grapes belong to the Vitacaee family group whose vines grow and produce dense fruit on their branches, grapes have health benefits for the body's metabolism. Wine has different types of variants, in this study there were 11 types of grape variants used consisting of Auxerrois grapes, Cabernet Franc, Cabernet Sauvignon, Chardonnay, Merlot, Muller Thurgau, Pinot Noir, Resling, Sauvignon Blanc, Syrah and Tempranillo. There are several ways to distinguish the types of grapes, one of which is by looking at the shape of the leaves of the tree. Grape leaves can be observed with the naked eye if people know and understand grape leaves, but if people who don't know or are still beginners don't understand grape leaves, the accuracy is not perfect because there are shapes of grape leaves that have a resemblance. To overcome this problem, an application is needed that makes it easier for the public to classify types of grapes automatically through a series of processing processes for the taste of grape leaves by recognizing the characteristics of the leaves such as the shape of the leaves. This study aims to classify types of grapes based on the shape of the leaves using the convolutional neural network (CNN) and K-nearest neighbor (KNN) methods to determine the types of grapes planted based on the shape of the leaves. The test results for the CNN and KNN methods were measured using a confusion matrix and obtained a result of 99% for CNN and 53% for KNN.