Measurement of Soluble Solids Content in Melon by Transmittance Spectroscopy

This research was focused on evaluating the use of Vis/NIR transmission spectroscopy for measuring soluble solids content (SSC) of intact melons. Two experiments were designed for two varieties of melons (Baipi, Hetao), which have different skin characters. The spectra were collected using a fiber spectrometer in the wavelength range of 350-1000 nm. Three different Spectral pre-processing methods of fist derivative, second derivative and Norris derivative spectra smoothing were used for establishing the partial least squares (PLS) models and principal component regression (PCR) models, which were determined the SSC in melons. Performance of different models was assessed in terms of correlation coefficient (r) of the calibration set, root mean square errors of calibration (RMSEC) and root mean square errors of prediction (RMSEP). The results showed that the model developed by the PLS method after first derivative pre-processing achieved the best performance, with high r (0.907 (Baipi); 0.921 (Hetao)), low RMSEC (0.631 degrees Brix (Baipi); 0.603 degrees Brix (Hetao)), low RMSEP (0.823 degrees Brix (Baipi); 0.796 degrees Brix (Hetao)) and small difference between the RMSEP and the RMSEC. The models based on Norris derivative smoothing spectra did not enhance the performance of calibration models.