Prediction of mechanical properties of blueberry using hyperspectral interactance imaging

The purpose of this investigation was to develop and validate a hyperspectral interactance imaging system to non-destructively estimate blueberry mechanical properties. Four texture profile analysis (TPA) and four puncture analysis (PA) parameters were predicted. A region growing based algorithm was used to segment the acquired interactance hypercubes and to assist in extracting mean spectra. Subsequently, the spectra were smoothed by Standard Normal Variate (SNV) and Savitzky-Golay first derivative (Der). Least squares support vector machines integrated with Monte Carlo uninformative variable elimination (MC-UVE) models were developed for mechanical parameters. Based on the MC-UVE selected wavelengths, the SNV model performed best for cohesiveness with R_p (R_c) value of 0.91 (0.91). The SNV models of springiness, resilience, max force strain and final force resulted in R_p (R_c) values of 0.84 (0.85), 0.86 (0.87), 0.65 (0.76) and 0.62 (0.72), respectively. Using Der spectra, the R_p (R_c) values were found to be 0.77 (0.86), 0.71 (0.73) and 0.58 (0.69) for hardness, maximum force and gradient, respectively. Generally, the overall performances of MC-UVE based models were similar to those with full spectra. The above results showed the potential of hyperspectral interactance imaging coupled with MC-UVE approach for predicting the mechanical properties of blueberry and the other small fruit.