Determination of honey floral origin using data fusion of electronic taste and smell sensing system

Abstract

Determination of honey classification based on floral origin requires expensive instrument and/or labor intensive expert personnel. In this study, electronic taste sensing and smell sensing systems were used to determine the floral origin of six varieties of honey from a local manufacturer. Classification using principal component analysis (PCA) on taste or smell data alone yields closely separated groups. Classification of fused data of taste and smell shows superiority to classification of taste or smell data alone. This study also shows that smaller number of (combined taste and smell) sensor is possible to properly classify honeys.