TY - JOUR
T1 - Prediction system for pH measurement on Brassica oleraceae (Red Cabbage) using machine learning regression
AU - Wulan Hastuti, Dian
AU - Harahap, Marlina
AU - Adila Ferdiansyah, Faizal
AU - Harmoko Saputro, Adhi
AU - Imawan, Cuk
N1 - Publisher Copyright:
© Published under licence by IOP Publishing Ltd.
PY - 2020/6/9
Y1 - 2020/6/9
N2 - pH is an important unit to represent the chemical condition of a liquid, solid substance, food nutrition, and microbial activity as well. pH value is also commonly used to detect the behaviour of chemical substances. Measurement of pH value can be shown by color changes of the substance based on the acidity condition from the measured environment. In this research, the colorimetric based on machine learning and pH value detected from the color information for point-of-care applications. For this investigation, we used the pH buffer solution and natural dyes derived from Brassica oleraceae (Red Cabbage) that shows colorimetric response gradually shifted from red to green along with the increasing of pH from 2.00 (acid) to 11.00 (alkaline). In this paper, we propose a method for predicting pH value based on Artificial Neural Network Regression (ANNR) and K-Nearest Neighbour Regression (KNNR) with RGB, HSV and LAB color space. As a result, the performance (99.83% ± 0.11) of this method could estimate the pH reasonably well for point-of-care application.
AB - pH is an important unit to represent the chemical condition of a liquid, solid substance, food nutrition, and microbial activity as well. pH value is also commonly used to detect the behaviour of chemical substances. Measurement of pH value can be shown by color changes of the substance based on the acidity condition from the measured environment. In this research, the colorimetric based on machine learning and pH value detected from the color information for point-of-care applications. For this investigation, we used the pH buffer solution and natural dyes derived from Brassica oleraceae (Red Cabbage) that shows colorimetric response gradually shifted from red to green along with the increasing of pH from 2.00 (acid) to 11.00 (alkaline). In this paper, we propose a method for predicting pH value based on Artificial Neural Network Regression (ANNR) and K-Nearest Neighbour Regression (KNNR) with RGB, HSV and LAB color space. As a result, the performance (99.83% ± 0.11) of this method could estimate the pH reasonably well for point-of-care application.
UR - http://www.scopus.com/inward/record.url?scp=85087045157&partnerID=8YFLogxK
U2 - 10.1088/1742-6596/1528/1/012050
DO - 10.1088/1742-6596/1528/1/012050
M3 - Conference article
AN - SCOPUS:85087045157
SN - 1742-6588
VL - 1528
JO - Journal of Physics: Conference Series
JF - Journal of Physics: Conference Series
IS - 1
M1 - 012050
T2 - 4th International Seminar on Sensors, Instrumentation, Measurement and Metrology, ISSIMM 2019
Y2 - 14 November 2019 through 14 November 2019
ER -