PENERAPAN FITUR WARNA UNTUK KLASIFIKASI JENIS BUAH ALPUKAT MENGGUNAKAN METODE TRESHOLDING DAN SUPPORT VECTOR MACHINE (SVM)

Abstract

One of the fruits that are commonly found in Indonesia is avocado (Perseaamericana mill). Avocado fruits have types that have different external features. However, not a few of these fruits also have the same outer shape so that it is difficult to distinguish from the eyes. In fact, not a few consider that a similar type of avocado is only one type, while in fact the fruit consists of several types. The purpose of this study is to identify or classify avocado fruit types based on color imagery features with tresholding segmentation methods and Support Vector Machine (SVM). Method The data used is avocado images totaling 180, which are classified into 3 types, alligator avocado, butter avocado, miki avocado. Taken features rgb average value color, rgb standard deviation, RGB skewness. The results of the trial showed that the Support Vector Machine method was able to classify avocado types well. In the data training process that uses 150 data and in the data testing process that uses 30 data. Each avocado used 10 testing data and obtained an alligator accuracy value of 80%, butter 90%, miki 50%, so that for the overall data it reached an accuracy value of 73.33%.