Articles
Interactance and reflectance near infrared spectroscopy for freshness evaluation of hen eggs
Article number
1213_97
Pages
637 – 642
Language
English
Abstract
Haugh units is an important index for evaluate freshness of hen eggs.
High score of Haugh units (≥60) from eggs means those are new fresh eggs.
This research is aimed to use near infrared spectroscopy for nondestructive prediction of eggs freshness by quantitative evaluation based on Haugh units.
Interactance mode (588-1091 nm) and reflectance mode (1000-2500 nm) of near infrared spectroscopy were investigated in this research.
Hen eggs from farm in Thailand were studied by storage at 25°C for 21 days.
Samples were taken for measurements at different days of storage (0, 4, 7, 10, 14, 18 and 21 days). A set of 247 samples (165 for calibration and 82 for prediction) was used for interactance mode and a set of 150 samples (102 for calibration and 48 for a prediction) was used for reflectance mode.
Calibration models were established and cross-validated using partial least squares regression (PLSR). The accuracies were considered by test in prediction groups.
The results showed that the interactance obtained better accuracy for prediction (correlation coefficient, R=0.91 and root mean square error prediction, RMSEP=5.64) when compared with reflectance mode (R=0.83 and RMSEP=7.11). In this study, the interactance near infrared spectroscopy is more suitable to use in application for freshness sorting of hen eggs.
High score of Haugh units (≥60) from eggs means those are new fresh eggs.
This research is aimed to use near infrared spectroscopy for nondestructive prediction of eggs freshness by quantitative evaluation based on Haugh units.
Interactance mode (588-1091 nm) and reflectance mode (1000-2500 nm) of near infrared spectroscopy were investigated in this research.
Hen eggs from farm in Thailand were studied by storage at 25°C for 21 days.
Samples were taken for measurements at different days of storage (0, 4, 7, 10, 14, 18 and 21 days). A set of 247 samples (165 for calibration and 82 for prediction) was used for interactance mode and a set of 150 samples (102 for calibration and 48 for a prediction) was used for reflectance mode.
Calibration models were established and cross-validated using partial least squares regression (PLSR). The accuracies were considered by test in prediction groups.
The results showed that the interactance obtained better accuracy for prediction (correlation coefficient, R=0.91 and root mean square error prediction, RMSEP=5.64) when compared with reflectance mode (R=0.83 and RMSEP=7.11). In this study, the interactance near infrared spectroscopy is more suitable to use in application for freshness sorting of hen eggs.
Authors
S. Suktanarak, S. Teerachaichayut, P. Jannok, P. Supprung
Keywords
nondestructive, prediction, Haugh units, sorting
Online Articles (97)