Menentukan Persentase Kematangan Buah Jeruk dengan Metode Matriks Co-occurrence

Abstract

Determining the ripeness level of citrus fruits in some communities is often performed manually, where this method is less effective. The lack of understanding in determining the maturity level of citrus fruits can be assisted through an automated system, namely image processing. The purpose of this study is to determine the percentage of citrus fruit maturity level. The method used is the co occurrence matrix. This method is used to obtain an overview of the frequency of pixel values in the co-occurrence matrix which wll be devided into 4 quadrants. The results of the co-occurrence matrix calculation are divided into two data, namely training data and testing data. The average results based on the co-occuence matrix which represents the maturity level are the raw and rotten categories tend to be in quadrant 1, because the colour is dark. Half-ripe and half-rotten categories tend to be in quadrant 1 and 4, because the colour is between dark and light. Ripe categories tend to be in quadrant 4, because of the light colour. The ripe category tends to be in quadrant 4, because of the light colour. The calculation results of the model equation show that the raw and mature categories have a maturity level between 10% and 100%, while the mature to rotten category has a maturity level between 100% and 185,5%. The accuracy calculation results show that the succes of the model in predicting maturity is 100%. The conclusion of this research is that the co-occurrence matrix method is able to identify the maturity of citrus fruit well.