Karakterisasi Mutu Fisik Biji Kakao Aksesi Papua Barat Menggunakan Pengolahan Citra Digital
Abstract
Karakterisasi biji kakao umumnya masih dilakukan secara manual yang
memiliki hasil kurang efisien. Dengan teknologi saat ini, karakterisasi biji kakao
bisa dilakukan dengan menggunakan pengolahan citra digital. Penelitian ini
bertujuan untuk melakukan karakterisasi parameter dimensi dan warna fisik biji
kakao kering menggunakan pengolahan citra digital, serta membangun metode
karakterisasi fisik biji kakao kering dengan pengolahan citra digital. Prosedur
penelitian meliputi persiapan sampel, pengukuran manual biji kakao, penentuan
parameter citra digital biji kakao, pengambilan citra, pembuatan program
pengolahan citra, serta perbandingan antara manual dan program. Hasil penelitian
menunjukkan persentase kesesuaian antara penggolongan secara program dengan
manual pada parameter area dengan massa, rasio height/width (H/W), rataan warna
red, rataan warna green, rataan warna blue, dan rataan penjumlah red, green dan
blue (RGB) dengan nilai light (L) masing-masing memiliki persentase sebesar
54,55%, 87,87%, 78,79%, 66,66%, 69,69%, dan 87,87%. Parameter citra digital
biji kakao yang berpotensi digunakan untuk karakterisasi biji kakao adalah width,
height, rataan warna red, rataan warna green, dan rataan warna blue. Berdasarkan
karakterisasi, CKR 12 tergolong aksesi yang unggul, baik berdasarkan mutu SNI
BSN 01-2323-2008 maupun dengan program pengolahan citra pada parameter
fisik. Cocoa bean characterization is generally still done manually, which has less
efficient results. With current technology, the characterization of cocoa beans can
be done using digital image processing. This study aims to characterize the physical
dimensions and color parameters of dried cocoa beans using digital image
processing, as well as develop a method for the physical characterization of dried
cocoa beans using digital image processing. The research procedure includes sample
preparation, manual measurement of cocoa beans, determination of digital image
parameters of cocoa beans, image capture, image processing program creation, and
comparison between manual and program. The results showed that the percentage
of conformity between program and manual classification on the parameters of area
to mass, height/width ratio (H/W), average red color, average green color, average
blue color, and average sum of red, green, and blue (RGB) with the value of light
(L) each had a percentage of 54.55%, 87.87%, 78.79%, 66.66%, 69.69%, and
87.87%. Digital image parameters of cocoa beans that can potentially be used for
cocoa bean characterization are width, height, red color average, green color
average, and blue color average. Based on characterization, CKR 12 is classified as
a superior accession, both based on the quality of SNI BSN 01-2323-2008 and with
image processing programs based on physical parameters.