Prediction of moisture, color and pH in cooked, pre-sliced turkey hams by NIR hyperspectral imaging system

Abstract

The investigation was conducted to develop a hyperspectral imaging system in the near infrared (NIR) region (900–1700nm) to predict the moisture content, pH and color in cooked, pre-sliced turkey hams. Hyperspectral images were acquired by scanning the ham slices (900–1700nm) originated from different quality grade of turkey hams. Spectral data were then extracted and analyzed using partial least-squares (PLSs) regression, as a multivariate calibration method, to reduce the high dimensionality of the data and to correlate the NIR reflectance spectra with quality attributes of the samples considered. Instead of using a wide range of spectra, the number of wavebands was reduced for more stable, comprehensive and faster model in the subsequent multispectral imaging system. From this point of view, important wavelengths were selected to improve the predictive power of the calibration models as well as to simplify the model by avoiding repetition of information or redundancies. With the help of PLS regression analysis, nine wavelengths (927, 944, 1004, 1058, 1108, 1212, 1259, 1362 and 1406nm) were selected as the optimum wavelengths for moisture prediction, eight wavelengths (927, 947, 1004, 1071, 1121, 1255, 1312 and 1641nm) for pH prediction and nine wavelengths (914, 931, 991, 1115, 1164, 1218, 1282, 1362 and 1638nm) were identified for color (a*) prediction. With the identified reduced number wavelengths, good coefficients of determination (R2) of 0.88, 0.81 and 0.74 with RMSECV of 2.51, 0.02 and 0.35 for moisture, pH and color, respectively, were achieved, reflecting reasonable accuracy and robustness of the models.