Non-destructive quality control detection of endogenous contaminations in walnuts using terahertz spectroscopic imaging

Abstract

The precise detection of foreign bodies in food is significant for guaranteeing food safety and quality. Terahertz (THz) spectroscopic imaging is a rapidly evolving technique with the advantages of non-destructive, non-ionization, and spectral fingerprinting characteristics, so it has attracted a lot of attention in the food industry. In this paper, the possibility of THz spectroscopic imaging in the detection of endogenous contaminations in a complex food matrix was explored. We first tried to detect endogenous foreign bodies in walnuts though comparing the typical absorption spectra of kernels and shells with different concentrations. The spectral classification could be realized by principal component analysis with an accuracy higher than 95%, which proved the feasibility of distinguishing shells from kernels, and then, THz imaging was applied to locate shells and kernels in tablets. The results of this study showed that THz spectroscopic imaging could discriminate shell contaminations from walnut kernels.