Pork is a perishable, nutrition and expensive food commodity , and its quality is related to bio-chemical changes is affect individual experience and preference. Pork meat must be grading and classification
for defined quality making information and pork is used appropriate processing. Liu et al.(2010) studied categorized pork quality using Gabor filter-based hyperspectral imaging technology, which combining spectral features from Gabor-filtered images and hyperspectral images(MS). They were found the average accuracy of pork quality level classification reached 84 ± 1%, while the average accuracy by 5 MS PCs was 72 ± 2%. However, results showed that different Gabor filters extracted different image texture features and furher yielded various classification results. Barbin et al.(2012) classified pork meat in tree quality (AMSA,2001) and using near-infrared hyperspectral imaging for measurement of quality. They were found optimal wavelengths to grading different pork and made pork class visualization from optimal wavelengths and PC score image. Near-infrared hyperspectral imaging was classification of different pork grades accurate, fast, non-destructive and chemical-free method.