VIS/NIR multivariate spectroscopic quality determination of blue crab meat

To obtain information needed to automate blue crab meat quality control, the application of Visible and Near Infrared Spectroscopy (VIS/NIR) in the quality determination of blue crab meat was investigated. Four quality parameters were determined: (1) color characteristics, (2) chemical composition, (3) thermal process effects and, (4) thermal properties. Spectral absorption data in the visible region was transformed into a color space to calculate standard CIE color values: lightness (L), red-green (a), and yellow-blue (b) by using this analytical tool. Chemical composition was determined by the proximate analysis for water, protein, carbohydrates, and fat. The thermal process studied was pasteurization. The thermal properties studied were thermal conductivity (k) and specific heat (Cp). Several statistical approaches such as analysis of variance (ANOVA), step wise multiple linear regression (SMLR), partial least squares (PLS), and principal components analysis (PCA) were applied to determine the appropriateness of the VIS/NIR method.;The results of this research were presented in four papers. It was demonstrated that it is possible to measure the color of crab meat using colorimetric methods. It was also determined that thermal process and animal harvest location are two important factors that contribute to discoloration. SMLR proved to be a better calibration method than PLS for determining crab meat composition. Water and protein contents were more accurately predicted by VIS/NIR spectroscopy than fat and carbohydrate contents. The error percentages in the calibrations were below 1.5% for all the characteristics. Color values were also determined accurately from the VIS region. The error percentage for determining color values was less than 5%. Selected wavelengths by the SMLR method were provided for the proximate compounds, water, protein, fat, and carbohydrates.;It was found that the longer the pasteurization, the lower the L and the a values. The b values were not affected by thermal treatment. By applying PCA to the whole VIS/NIR region of the electromagnetic spectrum, it was found that water content was a critical factor in thermal processing of crab meat and that there were effects of thermal treatment on the crab meat spectrum.;Specific heat and thermal conductivity were accurately predicted by VIS/NIR spectroscopy. The coefficient of determination for both thermal properties and meat types were above 0.9 and the standard errors of calibration and prediction were below 1.5%. A cross-validation was applied for the determination of thermal properties. All these findings represent a baseline for the automation of crab meat process, not only for quality screening, but also for instrument design and development.