1. Effect of fruit size on apple damage prediction.
- Author
-
Li, Bin, Wan, Xia, Zou, Ji-ping, and Lu, Ying-jun
- Subjects
PARTIAL least squares regression ,STANDARD deviations ,APPLES ,FRUIT - Abstract
The diverse size of apples can bring harmful effects on the quantitative assessment of their impact damage, so it is important to improve the quality of the apples damage sorting models by eliminating the effects of apples size. To solve the problem, the general models to satisfy the quantitative prediction of impact damage in apples with different fruit diameters were built by using hyperspectral imaging technology combined with mechanical parameters in the study. Firstly, the apples damage raw spectra were preprocessed by using three preprocessing methods, which were standard normal variable transformation, baseline calibration (Baseline) and convolutional smoothing, respectively, and the spectral data combined with the mechanical parameters (damage areas, average contact force, maximum force and damage volume) to build the partial least squares regression models. Then, the mixed fruit size group was selected by Kennard-Stone algorithm, and the prediction models of various mechanical parameters of single fruit size groups to mixed fruit size group apples were built. Finally, the general models for various mechanical parameters were built and reduced by using the Competitive Adaptive Re-weighting Algorithm. The results showed that the prediction effect of the models decreased when the large difference in fruit size between the modeling set and the prediction set. The general models predicted better after band screening, and the values of prediction set coefficients of determination (Rp
2 ) and the values of root mean square errors of prediction set for damage area, average contact force, maximum force, and damage volume are 0.8248 and 107.2266 mm2 , 0.7325 and 20.9469 N, 0.8355 and 32.862 N, 0.7657 and 942.6885 mm3 , respectively. Therefore, the general models can reduce the harmful effects of apple size and improve the accuracy of quantitative assessment of impact damage of apples during damage sorting process. [ABSTRACT FROM AUTHOR]- Published
- 2024
- Full Text
- View/download PDF