Skip to main content
SLU publication database (SLUpub)
Research article - Peer-reviewed, 2020

Use of near-infrared hyperspectral (NIR-HS) imaging to visualize and model the maturity of long-ripening hard cheeses

Priyashantha, Hasitha; Hojer, Annika; Saeden, Karin Hallin; Lundh, Ase; Johansson, Monika; Bernes, Gun; Geladi, Paul; Hetta, Marten

Abstract

Spectroscopic measurements and imaging have great potential in rapid prediction of cheese maturity, replacing existing subjective evaluation techniques. In this study, 209 long-ripening hard cheeses were evaluated using a hyperspectral camera and also sensory evaluated by a tasting panel. A total of 425 NIR hyperspectral (NIR-HS) images were obtained during ripening at 14, 16, 18, and 20 months, until final sensorial approval of the cheese. The spectral data were interpreted as possible compositional changes between scanning occasions. Regression modelling by partial least squares (PLS) was used to explain the relationship between average spectra and cheese maturity. The PLS model was evaluated with whole cheeses (average spectrum), but also pixelwise, producing prediction images. Analysis of the images showed an increasing homogeneity of the cheese over the time of storage and ripening. It also suggested that maturation begins at the center and spreads to the outer periphery of the cheese.

Keywords

cheese maturation; Principal component analysis; Partial least squares regression; Pixelwise image prediction

Published in

Journal of Food Engineering
2020, Volume: 264, article number: 109687