Hyperspectral imaging as process analysistechnology for inline applications – Laboratoryprecision meets high sampling accuracy

Hyperspectral Imaging (HSI) has found its way into
modern production lines as yet another camera technology that
combines the means of spectroscopy with computer vision. However,
while its imaging capabilities with its increased amount of
processed data are already subject to state of the art sorting machines
used for product integrity checks and foreign body removal,
the potential of doing real-time spectroscopy and quantitative
chemical analysis of product streams has up to now not
fully been harnessed for the production line. Respective industries
still rely solely on highly precise but statistically limited,
laboratory based, slow and invasive methods while HSI offers
the data needed to perform real-time and statistically representative
product sampling as add on. In this regard the potential
of HSI as process analysis technology (PAT) will be outlined in
this article presenting a first industrial field application based on
the HELIOS near infrared hyperspectral imaging camera. Results
from this up to now first commercialized application of this kind
in the potato industry, i.e. an inline dry matter detection for potatoes,
will be discussed.