From Structural Heterogeneity to Photon Transport Simulation: A Comprehensive Review of Optical Nondestructive Evaluation in Multilayered Fruit and Vegetable Matrices

Abstract

Optical nondestructive evaluation has become an indispensable tool for quality assessment in the food industry. Revealing light propagation mechanisms in multilayered fruit and vegetable tissues improves detection model accuracy and robustness. In fact, most natural and processed fruit and vegetable materials inherently exhibit layered structures, leading to inevitable deviations in optical signals interpretation and quality prediction. The complex structure and heterogeneous composition of fruit and vegetable matrices pose great challenges to light–tissue interaction analysis. This review summarizes the layer-specific differences in optical properties based on structural and compositional heterogeneity of fruit and vegetable matrices, using representative cases of other food matrices for comparison. It further addresses light propagation in multilayered fruit and vegetable tissues together with research on the optimization of detection parameters. Finally, a unified framework is proposed that integrates physicochemical properties, optical properties, and photon transport simulations to achieve mechanism-guided optimization for nondestructive detection. Notably, the advances of Monte Carlo (MC) simulation empowered by artificial intelligence (AI) are highlighted in denoising, accelerating computation, and improving prediction accuracy. AI-assisted MC simulation provides support for the efficient and high-precision prediction of light propagation in multilayer tissues, and lays a solid foundation for the progress of food photonics.