Description
According to the Specifications of Parmigiano Reggiano cheese, the rind percentage in grated cheese products should not exceed the 18% (w/w) threshold value (Specifications of Parmigiano Reggiano Cheese). In order to ensure product compliance to the quality standards, the present study has two main objectives: evaluating the potential of NIR-hyperspectral imaging (NIR-HSI) in the quantification of rind percentage, and estimating the effect of factors related to sample preparation and composition on the determination of this percentage. In the first step, hyperspectral images of grated cheese samples with varying levels of rind were acquired in the 1000 nm – 1650 nm NIR range. The hyperspectral images were converted into one-dimensional signals, named Common Space Hyperspectrogram (CSH), which are obtained by merging in sequence the frequency distribution curves of quantities derived from a Principal Component Analysis (PCA) model common to the whole image dataset (Calvini et al., 2016). The CSH signals were used to calculate a calibration model using Partial Least Squares (PLS) algorithm, in order to predict the rind amount of the corresponding samples. In the second step, fat content of the pulp and grater type were considered as potential factors influencing the spectral response for the quantification of rind percentage. Grated cheese samples were prepared considering all the possible combinations between three levels of rind amount (8%, 18% and 28%), two grater types, and two levels of fat content. The hyperspectral images were analysed by means of ANOVA Simultaneous Component Analysis (ASCA) in order to evaluate the influence of these factors and their interactions both on the spectral response and on the CSH signals.
Keywords: grated cheese, NIR-HSI, multivariate calibration, ASCA
Acknowledgements: First Author gratefully acknowledges receiving funding from Consorzio del Formaggio Parmigiano Reggiano (Italy).
REFERENCES
Calvini, R., Foca, G., Ulrici, A., 2016. Data dimensionality reduction and data fusion for fast characterization of green coffee samples using hyperspectral sensors. Anal. Bioanal. Chem. 408(26), 7351-7366. https://doi.org/10.1007/s00216-016-9713-7
Specifications of Parmigiano Reggiano Cheese. https://www.parmigianoreggiano.com/consortium/rules_regulation_2/default.aspx
| Consider for full paper in JNIRS | No, thank you |
|---|
