LAUSR

Abstract This study aim was to explore the laser backscattering imaging technique's capacity to model the curdling phase in cheese processing. To do so, three different formulas were studied by modifying solute concentration. Textural modifications to the matrix during curdling were characterised by viscosimetry and texture measurements depending on samples' liquid or solid state. This state changed by determining gelation to establish the limits for the liquid and solid phases. The process was also characterised by the imaging technique, which showed dependence on both solute concentration and enzymatic effect on both the previously observed phases. After following multivariate statistical procedures to reduce dimensionality, the imaging results revealed that solute concentration strongly influenced the variance that the imaging technique captured. It reduced the visibility of the phase change in the image parameters. After eliminating this influence, the evolution of the matrix across the liquid and solid phases was modelled. Data were divided into phases and used to successfully predict the matrix status in each phase by multivariate non-linear regression procedures. It was concluded that the laser backscattering imaging technique presented suitable properties to be used for non-destructive continuous curdling process monitoring during the cheese-making process.