Optical technology is an important and immerging technology for non-destructive and rapid detection of pork freshness.
This paper studied on the possibility of using multispectral imaging technique and scattering characteristics to predict the
freshness parameters of pork meat. The pork freshness parameters selected for prediction included total volatile basic
nitrogen (TVB-N), color parameters (L *, a *, b *), and pH value. Multispectral scattering images were obtained from
pork sample surface by a multispectral imaging system developed by ourselves; they were acquired at the selected
narrow wavebands whose center wavelengths were 517,550, 560, 580, 600, 760, 810 and 910nm. In order to extract
scattering characteristics from multispectral images at multiple wavelengths, a Lorentzian distribution (LD) function
with four parameters (a: scattering asymptotic value; b: scattering peak; c: scattering width; d: scattering slope) was used
to fit the scattering curves at the selected wavelengths. The results show that the multispectral imaging technique
combined with scattering characteristics is promising for predicting the freshness parameters of pork meat.