Abstract
A technique was developed to predict the freshness of packaged sliced chicken breast employing a nondestructive visible and short-wavelength near-infrared (SW-NIR) spectroscopy method. Spectra were recorded at 0, 7 and 14days using a camera, spectral filter (400-1000nm) and a halogen flood lighting system which were developed and calibrated for the purpose. Physicochemical, biochemical and microbiological properties such as moisture (x w), water activity (a w), pH, total volatile basic nitrogen (TVB-N), ATP breakdown compounds (K 1 values) and mesophilic bacteria (cfu g -1) were determined to predict freshness degradation. The spectra obtained were related to the storage time of the samples. The best wavelengths for modeling freshness were 413, 426, 449, 460, 473, 480, 499, 638, 942, 946, 967, 970 and 982nm. A linear correlation was found between the visible and SW-NIR spectroscopy and parameters such as microbiological counts, K 1 and T-VBN indexes. © 2010 Elsevier Ltd.
