Evaluation of pH of curry soup containing coconut milk by near infrared spectroscopy

The pH is the important parameters to characterize the food deterioration and an indicative of food spoilage. The aim of this research was to apply the near infrared (NIR) spectroscopy to evaluate pH of curry soup containing coconut milk. The soup samples from mixing tank, water content adjusted tank, UHT pipe and laminated containers in the production line were collected. There were also the pH adjusted samples where the curry was made from the same recipe but increasing placed time for 2, 4 and 6 hr after 0 hr. There are 73 samples in total. The sample was scanned with FT-NIR spectrometer. A prediction model for pH was established using NIR spectral data in conjunction with partial least squares regression, which was validated using leave one out cross validation and test set validation. After validated by unknown samples, the leave one out cross validation model showed better prediction performance. The best model developed using first derivative spectra in 9403.8-7498.3, 6102-5446.3 and 4605.4-4242.9 cm⁻¹ provided an coefficient of determination (r²), root mean square error of cross validation (RMSECV), bias and ratio of performance to interquartile (RPIQ) of 0.73, 0.28, 0.01 and 1.89. The model was usable for screening and some other "approximate" calibrations. The model could be improved for further development of robust model using more natural samples in evaluation of pH in the curry soup