Development and comparison of two handheld near-infrared spectrometer models for rapid Brix prediction in sugar beet

抄録

Existing techniques for evaluating the sugar content in sugar beet have considerable drawbacks, underscoring the need for more efficient methods. Here, we introduce a fast and minimally invasive method for measuring Brix in sugar beet with surface impurities using cost-effective handheld NIR spectrometers. Applicable in laboratories and potentially on-site, the method shows that spectral variations due to measurement locations minimally affect accuracy. The LinkSquare NIR spectrometer performed best with PLS regression on averaged intact spectra using smoothing and standard normal variate pre-treatment. The S-G1 spectrometer achieved superior results with PLS regression on the averaged peeled spectra without pre-treatment. These findings demonstrate the efficacy of handheld NIR spectrometers for sugar beet analysis and offer strategies for optimizing their accuracy and practicality.