Flesh texture quality is an important attribute for watermelons that can be used as a ripeness indicator for sorting fruits. A non-destructive method is employed here, based on vibrations response, to assess the consumer opinion about watermelon (Crimson Sweet) tissue. The responses of samples to vibration excitation were recorded by laser Doppler vibrometry (LDV). The amplitude and phase as two spectrums of frequency response function were extracted over a wide frequency range. Following non-destructive tests, the watermelons were sensory evaluated. The samples were graded in a range of ripeness by panel members in terms of texture acceptability. Stepwise multiple linear regression (SMLR) and partial least squares regression (PLSR) were applied to extracted vibration spectrums to construct prediction models of watermelon flesh quality. The results showed that performance of SMLR models on phase spectrum was better than others. The determination coefficients (R2) of the calibration and cross validation models were 0.9998 and 0.9986 respectively. This study demonstrated the feasibility of mentioned method for predicting the quality of watermelons in industrial grading systems.
Read full abstract