Food Texture Classification Using Magnetic Sensor and Principal Component Analysis

Abstract

Food texture is one of important characteristics which affect food sales. Designing and evaluating food textures require a food texture sensor. In this study, a novel food texture sensor is proposed. The food texture sensor has giant magnetoresistance (GMR) elements and an inductor as sensing elements. In food fracture, the GMR elements and the inductor measure force and induced voltage, respectively. The food texture sensor measure four foods in laboratory experiments and obtain their waveforms. A principal component analysis analyzes the waveforms of the force and induced voltage, and divided the food textures to clusters using principal components. Therefore, the force and induced voltage obtained by the sensor have the features of the waveforms reflecting the food textures.