Food texture evaluation using logistic regression model and magnetic food texture sensor

Abstract

A food texture evaluation method using a magnetic food texture sensor is proposed for the visualization of food texture. The food texture sensor measures two time-series waves, one of force and one of vibration, during fracture of a food sample. Twenty profiles were extracted from the two waves. The evaluation method selected the profiles to use in the logistic model and determined the coefficients of the model based on the results of sensory tests. Laboratory experiments confirmed that the logistic model evaluated the food textures as radar charts. In addition, the model can potentially evaluate the food textures of unknown foods.