Quantitative non-destructive characterization of apple fruit impact damage based on computation of the modulus of elasticity

Abstract

To address the issues of difficulty in measuring mechanical parameters and inconsistencies in damage grading standards caused by differences in the physiological structure and size of fruits, an elastic modulus-based quantitative damage prediction model was established. This laid the theoretical foundation for expanding the application scope of quantitative damage prediction models for fruits. Firstly, reflectance spectral data before and after apple damage are extracted, and mechanical parameters of apple collisions are collected. Then, the elastic modulus was used for linear fitting with other mechanical parameters. Finally, the Competitive Adaptive Re-weighting (CARS) algorithm was used to screen the characteristic wavelengths in the spectrum and establish prediction models for each mechanical parameter. The results indicate that there is a good linear relationship between the elastic modulus of apples and the damaged area, peak force, and damage depth. The prediction accuracy of the mechanical parameter model established based on CARS characteristic wavelength screening was improved. The R_p and RMSEP of their damage area, peak force, absorbed energy, damage depth, and elastic modulus change rate prediction models were 0.935 and 91.97 mm², 0.938 and 28.72 N, 0.896 and 0.24 J, 0.936 and 0.63 mm, and 0.920 and 4.07 %, respectively.