Nondestructive Detection of the Soluble Solids Content of Cherry Tomatoes Using Fiber Optic Spectroscopy and an SNV-CARS-GWO-SVR Model
Article
Figures
Metrics
Preview PDF
Reference
Related
Cited by
Materials
Abstract:
The soluble solids content (SSC) is a key parameter for evaluating the quality and maturity of cherry tomatoes. In this study, a fiber optic spectral transmission detection system was built to collect the raw spectral information of cherry tomatoes at different maturity levels. The SSCs of the samples were determined using physicochemical experiments, and the values were classified using the SPXY algorithm. Next, the raw spectra collected were preprocessed using standard normal variable (SNV) transformation. The successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS) algorithm were used for characteristic wavelength extraction. Finally, the Grey Wolf optimization (GWO) algorithm was applied to optimize the support vector regression (SVR) model to build an optimal model for predicting the SSC of cherry tomatoes. The results showed significant improvement of the coefficients of the calibration and prediction sets of the established prediction model based on SNV-preprocessed spectra. The SNV-CARS-GWO-SVR model was the best model for predicting the SSC of cherry tomatoes, with a root-mean-square-error of the prediction set (RMSEP) value of 0.28 and residual predictive deviation (RPD) value of 2.75. In summary, this independently developed fiber optic spectral transmission detection system fully achieved the nondestructive detection of the SSC of cherry tomatoes, thus providing a new method for the rapid and nondestructive online detection of SSCs of cherry tomatoes at different maturity levels.
