Smart horticulture era: advances in hyperspectral imaging for intelligent non-destructive quality assessment of fruits and vegetables

Abstract

The rising consumer demand for safe, high-quality produce has spurred interest in efficient, non-destructive testing technologies. Hyperspectral imaging (HSI), integrating spectroscopy and imaging techniques, offers significant potential for rapid, precise, and non-destructive quality assessment of fruits and vegetables. HSI effectively predicts internal quality parameters, achieving high accuracy in firmness and SSC assessments for various produce, utilizing advanced machine learning and deep learning algorithms. It significantly improves the accuracy of maturity classification and external defect detection compared to traditional methods. Notably, HSI shows great promise for rapid pesticide residue detection. Nevertheless, HSI faces challenges such as slow data processing, high dimensionality, limited model generalization, and substantial equipment costs. Future research should emphasize enhancing algorithm efficiency, improving hardware affordability, and expanding applicability across diverse produce types to facilitate broader industry adoption.