Transmittance spectra and acoustic properties of durians with different ripening: An exploration of application for complex-structured and large-sized fruit

The complex structure and large size of durians pose limitations on the application of ripeness detection using spectroscopic and acoustic technology. In order to explore the feasibility of online detection of durian ripeness, transmittance spectra in the wavelength range of 400–1175 nm and acoustic signals of durian samples with different ripeness were collected. The effects of durian physical properties on the transmittance spectra and acoustic properties at various ripening were analyzed. The differences in the light and sound propagation path were found, in the detection of durians with complex structure and large size. The experimental results indicated that average transmittance spectra of durians increased with advancing ripeness. Furthermore, circumference correction was used to reduce the effect of durian size on the spectra. After correction, the average correlations of the intensity at 860 nm and the intensity ratio I860/I670 with ripeness increased to 0.794 and 0.761, respectively. Acoustic parameters were optimized based on correlation with ripeness. The preferred acoustic parameters, i.e., Elastic Index and average grey value decreased with increasing ripeness, with average correlations of −0.792 and −0.775, respectively. Based on the selected parameters, probability of success in the classification of durian ripeness had been achieved using support vector machine (SVM), with an accuracy of 80 %. For the complex-structured and large-sized durian, pursuit of eliminating the impact of physical properties on its transmittance spectra and acoustic properties was of great significance. The results undoubtedly cast light on the ripeness detection of durian samples based on transmittance spectra and acoustic properties. [Display omitted] •Effect of durian physical properties on the spectra and acoustic properties was analyzed.•Correction was used to minimize the impact of durian size on the spectra.•Acoustic parameters of higher correlation with ripeness were selected.•Satisfactory ripeness classification was achieved with selected parameters using PCA-SVM.