Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Abstract. Hyperspectral imaging combined with chemometric techniques were developed for rapid quantification of cheaper starches (cassava starch) adulterated illegally in potato starch. Cassava starch was mixed into potato starch powder, with the doping levels of 1, 3, 5, 10, 15, 25, 35, 45, 55, 65, and 75% (w/w). Hyperspectral images of the adulterated samples with a spectral range of 936-1700 nm were collected. The acquired spectra were preprocessed by standard normal variate (SNV), moving average smoothing (MAS), Savitzky-Golay first derivative (SG1D), and Savitzky-Golay second derivative (SG2D) before modeling. Partial least squares regression (PLSR) was performed for quantitative analysis of adulteration proportion of cassava starch in potato starch. Successive projections algorithm (SPA) were used to identify the most important wavelengths to develop effective and simplified models. The optimal result was generated by the SNV-SPA-PLSR model, with a determination coefficient of prediction (R_p²) of 0.9946 and a root mean square error of prediction (RMSEP) of 1.9992%. Visualization maps were generated by applying multispectral PLSR models on the spectra of each pixel of powder samples images. The results indicated hyperspectral imaging technique has the potential to be an objective and non-destructive method to quantify cassava starch adulterated in potato starch powder.