Mechanism guided near-infrared determination of soluble solids content in whole European plum based on optical...

Soluble solids content (SSC) is a key indicator of sweetness and quality in European plum, yet rapid and non-destructive detection methods remain limited. This study aimed to develop an optical property-guided near-infrared (NIR) detection method for SSC in whole European plum. An integrating sphere system combined with the inverse adding-doubling (IAD) algorithm was used to determine the absorption coefficient (μ_a), reduced scattering coefficient (μ_s'), and anisotropy factor (g) of plum peel and flesh. The relationships between optical properties and SSC were analyzed, and optical penetration depths were calculated. Voxel-based Monte Carlo simulations were further performed to investigate photon transport in plum tissues and to optimize the detection configuration. Results showed that SSC variation was mainly associated with the absorption coefficient of the flesh. Light in the range of 900-1350 nm could penetrate the peel and reach the flesh tissue. Diffuse reflectance was identified as an effective detection mode, and the integrating sphere enabled a balance between sampling depth and signal intensity. The informative spectral regions were further refined to 900-950 nm and 1010-1125 nm based on penetration depth and detector sampling characteristics. Based on the optimized detection scheme, NIR spectra were collected and separately preprocessed using SG and MSC before modeling. Prediction models were developed using PLSR, SVR, and LSSVM. The results of the validation experiments showed that the MSC-LSSVM model achieved the best performance (R² = 0.887, RMSE = 0.024, and RPD = 2.946), indicating reliable predictive ability for SSC. This study provides a mechanistic and practical basis for non-destructive SSC detection in European plum and supports the development of rapid optical sensing methods.