PENENTUAN LAMA SANGRAI KOPI BERDASARKAN VARIASI DERAJAT SANGRAI MENGGUNAKAN MODEL WARNA RGB PADA PENGOLAHAN CITRA DIGITAL (DIGITAL IMAGE PROCESSING)

Abstract

Roasting is one of the important stages in coffee processing to forming its aroma and taste. The purpose of this study was to determine the level of relationships between RGB value from sample image with the RGB value from roasting degrees standard image and predict the duration of roasting based on the similarity of RGB value between both samples and roasting degrees standard. The samples of roasted coffee were taken in various temperature (195 o C – 250 o C). Images of those samples were taken using CCD camera and IC Capture software. The images then were evaluated using Sharp Develop program to obtain RGB values. There were very strong relationships between RGB value Arabica samples with roasting degrees standard of roast temperature 205°C, 225°C, 250°C, either with robusta samples, at temperature 195°C, 200°C, 205°C , 225°C, 250°C and 275°C. Based on analysis, the G value produced the best relationship, so the formula to predict the duration of roasting build with this value. Using linier regretion all of the duration to achieved various roasting degrees of Arabica and robusta can be predicted at every roast temperature from the lowest degrees (cinnamon roast) to highest degrees (Italian roast) at every roast temperature. The Result were: Robusta 200°C (101.7 minutes, 158.2 minutes); Robusta 225°C (56.9 minutes, 85.2 minutes); Robusta 250°C (30.2 minutes, 41.3 minutes); Arabica 205°C (31.6 minutes, 44.7 minutes); Arabica 225°C (20.7 minutes, 28.9 minutes); and Arabica 250°C (21 minutes, 29.6 minutes).