Evaluation of Food Freshness and Locality by Odor Sensor

Abstract

The aim of this study was to investigate whether food freshness and locality can be classified using a food evaluation system consisting four SnO2-semiconductor gas sensors and a solid phase column, into which collecting aroma materials. The temperature of sensors was periodically changed to be in unsteady state and thus, the sensor information was increased. The parameters (in quefrency band) were extracted from sensor information using cepstrum analysis that enable to separate superimposed information on sinusoidal wave. The quefrency was used as parameters for principal component and discriminant analyses (PCA and DCA) to detect food freshness and food localities. We used three kinds of strawberries, people can perceive its odors, passed from one to three days after harvest, and kelps and Ceylon tea, people are hardly to perceive its odor, corrected from five areas as sample. Then, the deterioration of strawberries and localities of kelps and Ceylon teas were visually evaluated using the numerical analyses. While the deteriorations were classified using PCA or DCA, the localities were classified only by DCA. The findings indicate that, although odorant intensity influenced the method detecting food quality, the quefrency obtained from odorant information using cepstrum analysis were available to detect the difference in the freshness and the localities of foods.