Detection of Bacterial Contamination of Pork using Hyperspectral Scattering Technique

The hyperspectral scattering technique was investigated for determining the total viable counts of pork in the research. Pork samples were purchased from a local supermarket and stored at 4°C for 14 days. The hyperspectral images of pork samples were acquired in the spectral range of 400nm~1100nm every 24 hours, in parallel the total viable counts of pork were obtained by traditional method. Both the 4-parameter Lorentzian function and Gompertz function were applied to fit the scattering profiles of pork and the fitting results were good in 570 nm~780 nm, 460 nm~960 nm respectively. Results showed that the individual Lorentzian parameters a, b, c and Gompertz parameters a, s, e can give good predictions for TVC of pork, and would be more precise if integrated respectively. The best result can be acquired using the integrated Gompertz parameter with RCV, SECV of 0.929 and 0.781. The research provides a promising tool for rapid non-destructive detection of bacterial contamination of meat.