Geladi, Paul
- Department of Forest Biomaterials and Technology, Swedish University of Agricultural Sciences
Abstract
Adulteration of food products remains a food security risk. Spices are food components with a high value per unit mass due to their desired flavour attributes and are therefore economically worthwhile targets for adulteration. Vibrational spectroscopy techniques could be ideal to detect adulterants due to benefits such as speed of analysis. Near infrared (NIR) and mid-infrared (mid-IR) spectra were used to quantify the amount of adulterant (buckwheat or millet) in ground black pepper. NIR spectra used for calibration were the average spectra of individual hyperspectral images recorded with a spatial resolution of 300 mu m x 300 mu m from 1000 nm to 2500 nm at 6.3 nm intervals; hyperspectral images were collected to cope with the heterogeneity of the samples. The calibrations calculated, using the averaged NIR spectra, were more accurate than those calculated from mid-IR spectra. This was a direct result of sample heterogeneity and the insufficient sampling area in the mid-IR measurements. NIR-based calibrations were suitable for process control when the appropriate spectral data pre-treatment was used: standard normal variate followed by first derivative pre-processing of the 1100-2500 nm spectral range resulted in a root mean square error of prediction equal to 2.7% w/w and a ratio of standard error of prediction to standard deviation (validation set) of 11.1.
Keywords
adulteration; black pepper; spice; image prediction; regression modelling; spectral pre-processing
Published in
Journal of Near Infrared Spectroscopy
2012, volume: 20, number: 5, pages: 521-528
Publisher: N I R PUBLICATIONS
SLU Authors
Global goals (SDG)
UKÄ Subject classification
Other Chemistry Topics
Publication identifier
- DOI: https://doi.org/10.1255/jnirs.1008
Permanent link to this page (URI)
https://res.slu.se/id/publ/45069