Karakterisasi Kematangan Buah Kopi Berdasarkan Warna Kulit Kopi Menggunakan Histogram dan Momen Warna

Main Article Content

Hendri Syahputra
Fitri Arnia
Khairul Munadi

Keywords

Abstract

Conventionally, the coffee maturity level is determined by observing the fruit colour, and it is done manually. This approach may result in inconsistency in colour classification. Thus, an automatic colour classification method based on colour of coffee maturity level is required. This paper presents the characterization of coffee maturity level based on two colour features: colour histogram and colour moment. Characterization of coffee maturity level was grouped into four class: green for unripe coffee, greenish-yellow for half ripe coffee, red for ripe coffee, and dark red for too ripe coffee. The purpose of the research is to determine the colour features that can characterize the coffee maturity level based on computer simulation in extracting and calculating the statistical values of the colour histogram and colour moments. It turned out from 200 coffee images that the statistical values of colour histogram are more suitable for characterising the coffee maturity. The kurtosis values of hue histogram for each maturity level of coffee were different: kurtosis value of unripe coffee was 17.2-28.3, those of half ripe coffee, ripe coffee and too ripe coffee were 29.2-31.4, 32.7-83.5, and more than 84.2 respectively.

.

Keywords : colour histogram kurtosis, colour moment, image processing.


Abstrak

Secara tradisional, tingkat kematangan buah kopi ditentukan dari warna kulitnya yang dikelompokan secara manual. Cara ini menghasilkan pengelompokan warna yang kurang konsisten, sehingga diperlukan sebuah metode otomatis pengelompokan buah kopi berdasarkan warna dari tingkat kematangannya. Penelitian ini memaparkan hasil karakterisasi kematangan buah kopi arabika menggunakan dua fitur warna citra, yaitu histogram dan momen warna. Karakterisasi kematangan dibagi menjadi empat kelompok: hijau untuk kopi muda, hijau kekuningan untuk kopi setengah masak, merah untuk kopi masak, dan merah tua untuk kopi tua. Tujuan penelitian ini adalah menentukan fitur warna yang dapat mewakili karakter kematangan buah kopi dengan melakukan simulasi komputer untuk mengekstrak dan menghitung nilai statistik dari histogram warna dan nilai momen warna dari empat kelompok buah kopi.  Hasil penelitian menggunakan 200 citra kopi menunjukkan bahwa nilai statistik dari histogram warna lebih menggambarkan karakter kematangan buah kopi, dibandingkan dengan momen warna. Nilai kurtosis dari histogram hue memiliki nilai berbeda untuk setiap kategori kematangan buah kopi: kopi muda memiliki nilai kurtosis 17.2-28.3, kopi setengah masak 29.2-31.4, kopi masak 32.7-83.5dan kopi tua lebih dari 84.2.

  

Kata Kunci : kurtosis histogram warna, momen warna, pengolahan citra.

References

P. Rahardjo, “Kopi Panduan Budidaya dan Pengolahan Kopi Arabika dan Robusta,” Jakarta: Penerbit Penebar Swadaya, 2012.

[2] A.S, Somantri, “Teknologi pengolahan citra digital untuk identifikasi mutu fisik produk tanaman perkebunan,” Warta Penelitian dan Pengembangan Tanaman Industri, Pusat Penelitian dan Pengembangan Perkebunan, 2009.

[3] B. M. Ayitenfsu, “Method of Coffee Bean Defect Detection,” International Journal of Engineering Research & Technology (IJERT), vol. 3, no. 2, pp. 2355-2357, Feb, 2014.

[4] F. Faridah, G. O. F. Parikesit, F. Ferdiansjah “Coffee Bean Grade Determination Based on Image Parameter,” TELKOMNIKA, vol.9, no.3, pp. 547-554, Dec, 2011.

[5] B. Turi, G. Abebe, G. Goro, “Classification of Ethiopian Coffee Beans Using Imaging Techniques,” East African Journal of Sciences,
vol. 7, no. 1, pp. 1-10, Jan, 2013.

[6] E. Carrillo dan A. A. Peñaloza, “Artificial Vision to assure Coffee-Excelso Beans quality,” The 2009 Euro American Conference on Telematics and Information Systems: New Opportunities to increase Digital Citizenship, Prague, 2009.

[7] R. Calvini, A. Ulrici, J. Amigo, “Practical comparison of sparse methods for classification of Arabica and Robusta coffee species using near Infra-red hyperspectral imaging,” Chemometrics and Intelligent Laboratory Systems, vol. 146, pp. 503-511, Aug, 2015.

[8] E. M. Oliveira, et.al, “A computer vision system for coffee beans classification based on computational intelligence techniques,” Journal of Food engineering, vol 171, pp. 22-27, Feb, 2016.

[9] J. C. Caban. (2010, Sep.). Introduction to Image Statistics. [Online]. Available https://www.csee.umbc.edu/~caban1/Fall2010/CMSC691//Schedule_files/Docs/08-ImageStatistics.pdf

[10] N. Keen, “Color Moments,” unpublished.