Using RGB Imaging, Optimized Three-Band Spectral Indices, and a Decision Tree Model to Assess Orange Fruit Quality

Point samples and laboratory testing have historically been used to evaluate fruit quality criteria. Although this method is precise, it slow, expensive, destructive, making unsuitable for large-scale monitoring of these parameters. The main objective research was develop a non-invasive protocol by combining color RGB indices (CIs) previously published newly developed three-band spectral reflectance (SRIs) with decision tree (DT) model the parameters navel orange. These were brightness (L*), red–green (a*), blue–yellow (b*), chlorophyll meter (Chlm), total soluble solids (TSS), TSS/acid ratio. characteristics orange measured at various stages ripening. outcomes demonstrated that levels ripening, parameters, imaging indices, SRIs differed. based on wavelengths blue, green, red, red-edge, NIR are most effective estimating six in study. For example, NDI574,592,724, NDI572,584,724, NDI574,722,590 had largest R2 value (0.90) L*, whereas NDI526,664,700 NDI524,700,664 exhibited highest (0.97) a*. Moreover, integrating CIs DT has provided potentially useful tool accurate measurement studied instance, DT-SRIs-CIs-30 performed better terms measuring a* using 30 indices. 0.98 RMSE = 1.121 cross-validation, while 0.964 2.604 test set. Otherwise, fusion five DT-SRIs-CIs-5 precise recognizing b* (R2 0.957 0.929, 1.713 3.309 cross-validation set, respectively). Overall, work proves different proximal sensing systems such as via may be considered reliable instrument evaluating fruits.