Universal Journal of Agricultural Research 10(4): 397-404, 2022 http://www.hrpub.org
DOI: 10.13189/ujar.2022.100409
Characterization and Classification of Citrus reticulata
var. Keprok Batu 55 Using Image Processing
and Artificial Intelligence
Aimmatuz Zakiyyah
1
, Zainuri Hanif
3
, Dina Wahyu Indriani
1,2
, Zaqlul Iqbal
2
, Retno Damayanti
2
,
Dimas Firmanda Al Riza
1,2,*
1
Bioprocess Engineering Study Program, Faculty of Agricultural Technology, Universitas Brawijaya, Jl. Veteran, Malang, Indonesia
2
Agricultural Engineering Department, Faculty of Agricultural Technology, Universitas Brawijaya, Jl. Veteran, Malang, Indonesia
3
Research Center for Behavioral and Circular Economics, National Research and Innovation Agency, Jalan Jend, Gatot Subroto, No.10
South Jakarta 12710, Indonesia
Received April 16, 2022; Revised July 7, 2022; Accepted July 20, 2022
Cite This Paper in the Following Citation Styles
(a): [1] Aimmatuz Zakiyyah, Zainuri Hanif, Dina Wahyu Indriani, Zaqlul Iqbal, Retno Damayanti, Dimas Firmanda Al
Riza , "Characterization and Classification of Citrus reticulata var. Keprok Batu 55 Using Image Processing and
Artificial Intelligence," Universal Journal of Agricultural Research, Vol. 10, No. 4, pp. 397 - 404, 2022. DOI:
10.13189/ujar.2022.100409.
(b): Aimmatuz Zakiyyah, Zainuri Hanif, Dina Wahyu Indriani, Zaqlul Iqbal, Retno Damayanti, Dimas Firmanda Al
Riza (2022). Characterization and Classification of Citrus reticulata var. Keprok Batu 55 Using Image Processing and
Artificial Intelligence. Universal Journal of Agricultural Research, 10(4), 397 - 404. DOI: 10.13189/ujar.2022.100409.
Copyright©2022 by authors, all rights reserved. Authors agree that this article remains permanently open access under the
terms of the Creative Commons Attribution License 4.0 International License
Abstract Citrus reticulata var. Keprok Batu 55 is one
of the superior varieties of citrus originating from Batu
City, East Java, which has a slightly sour-sweet taste with a
sweetness level of 10-12
o
brix. Prediction of citrus maturity
as a monitoring activity for pre- and post-harvest quality
management is still done manually, whereas human
judgment of the maturity level is subjective. One
alternative to increase the monitoring productivity is the
development of a portable system with image processing
and destructive measurements of physico-chemical
properties such as hardness, brix, and pH. This study aims
to develop an image-based classification model and
characterize the quality parameters of citrus. Measurement
of maturity on Citrus reticulata var. Keprok Batu 55 has
been carried out for image analysis with color index (RGB,
L*a*b. and HSV). The image of citrus will be taken with
the camera, which will later be taken partially (cropping)
on the skin, which will then extract the color characteristics
and calculate the level of color content from RGB and then
converted it to HSV. A sufficient number of images with
various conditions are needed to train the artificial
intelligent model so that it can perform segmentation,
calculation, and grade classification. A prediction model
then was developed using color features and several
machine learning modeling approaches.
Keywords Brix, Classification, Hardness, Machine
Learning, Maturity, pH
1. Introduction
Citrus are sub-tropical fruits which are widely cultivated
in many countries even in tropical countries such as
Indonesia. Citrus can be grown in the lowlands to the
highlands depending on the varieties. There are more than
250 citrus varieties cultivated in Indonesia. One of the
Indonesia’s superior qualities of citruses is the Citrus
reticulata var. Keprok Batu 55. This citrus is popular due
to its juiciness, high vitamin-C content (32.27 mg/100 g)
and sweet taste with a sweetness level of 10-12
o
brix [1].
In citrus cultivation, monitoring of plant conditions and
fruit growth is needed for various things, such as prediction
of peak harvest, yield prediction, implementation of
precision agriculture in the fertigation process, and so on.
The level of maturity is one of the factors that affect
consumer preference. Maturity can be determined by