KLASIFIKASI KEMATANGAN BUAH ALPUKAT MENTEGA BERDASARKAN FITUR WARNA MENGGUNAKAN SUPPORT VECTOR MACHINE

Abstract

The ripeness level of the avocado greatly affects the shelf life of the fruit and also determines the taste of the fruit. Determining the proper ripeness of a fruit will play an important role in increasing the nutritional value and ripeness of avocado affects the quality of avocado oil. Classification of fruit maturity manually has many limitations because it is influenced by human subjectivity so that the application of digital image processing needs to be used.This study aims to classify avocado fruit ripeness which is divided into three categories, namely Unripe, Ripe, and Rotten. This research is intended to facilitate the classification of ripeness in fruit through its color. This study used avocado fruit as an experimental material which was carried out using the SVM (Support Vector Machine) method. The colors of the various avocados and in different positions and conditions of light contrast are used as data to classify the types of avocados. Data collection was taken from the selected Keagle Dataset of 150 data. The data is grouped into two data sets, namely dataset-1 (120 training data and 30 test data) and dataset-2 (90 training data and 60 test data). The image will be converted to grayscale to facilitate the classification process and look for HSV (Hue, Saturation, Value) and RGB (Red, Green, Blue) values to classify the fruit. The results from dataset-1 has an accuracy rate of 86%, 90% precision, and 86% Recall, while dataset-2 has an accuracy rate of 85%, 84% precision, and 85% Recall.