Fei Ma1, Bin Zhang2, Wu Wang1, Peijun Li1, Xiangli Niu1, Conggui Chen1, Lei Zheng1. 1. School of Food Science and Engineering, Hefei University of Technology, Hefei, Anhui Province, China. 2. Department of Biology and Food Engineering, Bengbu College, Bengbu, Anhui Province, China.
Abstract
BACKGROUND: The traditional detection methods for moisture content (MC) and water-holding capacity (WHC) in cooked pork sausages (CPS) are destructive, time consuming, require skilled personnel and are not suitable for online industry applications. The goal of this work was to explore the potential of multispectral imaging (MSI) in combination with multivariate analysis for the identification of MC and WHC in CPS. RESULTS: Spectra and textures of 156 CPS treated by six salt concentrations (0-2.5%) were analyzed using different calibration models to find the most optimal results of predicting MC and WHC in CPS. By using the fused data of spectra and textures, partial least squares regression models performed well for determining the MC and WHC, with a correlation coefficient (r) of 0.949 and 0.832, respectively. Additionally, their spatial distribution in CPS could be visualized via applying prediction equations to transfer each pixel in the image. CONCLUSION: Results of satisfactory detection and visualization of the MC and WHC showed that MSI has the potential to serve as a rapid and non-destructive method for use in sausage industry.
BACKGROUND: The traditional detection methods for moisture content (MC) and water-holding capacity (WHC) in cooked pork sausages (CPS) are destructive, time consuming, require skilled personnel and are not suitable for online industry applications. The goal of this work was to explore the potential of multispectral imaging (MSI) in combination with multivariate analysis for the identification of MC and WHC in CPS. RESULTS: Spectra and textures of 156 CPS treated by six salt concentrations (0-2.5%) were analyzed using different calibration models to find the most optimal results of predicting MC and WHC in CPS. By using the fused data of spectra and textures, partial least squares regression models performed well for determining the MC and WHC, with a correlation coefficient (r) of 0.949 and 0.832, respectively. Additionally, their spatial distribution in CPS could be visualized via applying prediction equations to transfer each pixel in the image. CONCLUSION: Results of satisfactory detection and visualization of the MC and WHC showed that MSI has the potential to serve as a rapid and non-destructive method for use in sausage industry.