ScholarWorks@경상국립대학교

초록

Purpose: The addition of harmful food additives (melamine, maleic anhydride, and benzoyl peroxide) to powdered foods (milk, starch, and wheat powder, respectively) has been widely recognized as a critical issue because of its negative effects on human health. To quantitatively detect additives in powdered food, a short-wave infrared (SWIR) hyperspectral imaging (HSI) system integrated with detection models was?utilized. Methods: The used?SWIR HSI system consists of a SWIR HSI camera module (1000?2500 nm) with a mercury cadmium telluride imaging unit, halogen lamps, aluminum sample holders, a control computer, and a stepper motor for sample translation. For classification between food additives and powdered food, partial least squares regression and analysis of variance-optimized models were utilized according to the additive concentration. Results: Hyperspectral detection images of food additives were obtained using binary images. The resultant hyperspectral images demonstrated that melamine, maleic anhydride, and benzoyl peroxide could be detected in the powder mixture. The resultant detection images were described as binary images, and the used SWIR HSI system showed high determinant coefficients (>0.89) between predicted and actual values. Conclusion: The SWIR HIS system and optimized models showed a high potential for discriminating food additives from powdered food. ? 2022, The Author(s), under exclusive licence to The Korean Society for Agricultural Machinery.

키워드