A Study on Classification of Food Texture with Recurrent Neural Network

This study constructs a food texture evaluation system using a food texture sensor having sensor elements of 2 types. Characteristics of food are digitized by using the food texture sensor in imitation of the structure of the human tooth. Classification of foods is carried out by the recurrent neural network. The recurrent neural network receives the time-series outputs from the food texture sensor, and outputs classification signals. In the experiment, 3 kinds of food are classified by the recurrent neural network.