This article discusses the first known application of the non-contact piezoelectric ultrasound system for physical property measurement of food with specific application to cheese. A non-contact ultrasound parameter called the relative attenuation and the ultrasound velocity were measured and correlated with the physical properties of Cheddar cheese (such as failure strain, failure stress, Young's modulus, and toughness) using a multi-linear neural network analysis. When the relative attenuation, velocity, and cheese types were used as inputs to the neural network, the squared correlation values for the prediction of failure strain, failure stress, secant modulus, and toughness of Cheddar cheese were 0.836, 0.947, 0.625, and 0.956, respectively. Results indicated that the non-contact ultrasound technique has excellent potential for rapid and automatic quality measurement of the physical properties of Cheddar cheese.
展开▼
