Spectral sensing in the near-infrared and chemometrics for flour quality multi-analysis

Abstract

Measuring flour quality is crucial for consistently achieving high-quality results in baking and food production. It empowers bakers and food manufacturers to make decisions regarding ingredient selection, adjust recipes for optimal performance, and meet consumer expectations. The advantage of optical spectroscopy in assessing flour quality lies in its non-destructive nature, speed, and the capability to analyze multiple components simultaneously without requiring any sample preparation.We conducted an examination of 46 commercially available flours of various types, including cereals, legumes, tubers, and others. Consequently, their nutraceutical composition varied. For spectroscopic measurements in the 850-1700 nm range, we employed the SpectraPod TM pocket-size spectral sensor. Subsequently, we utilized chemometrics to integrate the measured spectra with the nutraceutical data provided on the flour packaging. Chemometric processing of spectroscopic data enabled us to differentiate the samples based on low and high concentrations of carbohydrates and proteins, with cutoff levels set at 30 g/hg for carbohydrates and 16 g/hg for proteins, achieving an accuracy of 93.5% and 82.6%, respectively.