Canola quality by handheld NIR spectrometer

Handheld NIR spectrometers can be used to predict some of the most important compounds that define canola quality (high oil and protein contents, low glucosinolate and chlorophyll contents, and fatty acid composition). It is important to test models on true external verification sets instead of relying only on cross-validation for model stability. For handheld NIR spectrometers, the compact design necessarily has small sample windows, necessitating multiple measurements in intact seeds to reduce sampling error. Ground/milled seeds or flour samples have lesser sampling effect. It is important to understand how the limited wavelength range of some handheld NIR spectrometers makes NIR spectrometers unsuitable for the prediction of certain quality parameters.