Study on the Quantitative Assessment of Impact Damage of Yellow Peaches Using the Combined Hyperspectral Technology and Mechanical Parameters
Table 1
PLSR model prediction results of spectra after preprocessing with different methods.
Spectral pretreat type
Parameter
PCs
Modeling set
Prediction set
RMSEC
RC
RMSEP
RP
RPD
Raw
DA (mm2)
8
83.515
0.925
88.664
0.923
2.583
AE (J)
8
1.351
0.823
1.433
0.815
1.432
MCF (N)
9
64.096
0.904
71.455
0.879
1.969
MS (MPa)
8
0.162
0.699
0.147
0.642
0.954
SNV
DA (mm2)
7
94.885
0.903
102.021
0.891
2.111
AE (J)
13
1.271
0.849
1.337
0.838
1.521
MCF (N)
11
62.520
0.909
70.319
0.907
2.225
MS (MPa)
6
0.164
0.695
0.145
0.664
1.121
MSC
DA (mm2)
7
93.509
0.905
105.599
0.878
1.923
AE (J)
13
1.254
0.857
1.317
0.829
1.512
MCF (N)
11
61.955
0.911
69.961
0.901
2.224
MS (MPa)
7
0.162
0.695
0.148
0.641
0.966
SG
DA (mm2)
9
82.974
0.926
91.779
0.920
2.559
AE (J)
9
1.306
0.836
1.498
0.799
1.340
MCF (N)
9
65.504
0.900
70.723
0.881
1.980
MS (MPa)
8
0.161
0.696
0.155
0.614
0.887
DA: damaged area; AE: absorbed energy; MCF: maximum contact force; MS: maximum stress. PCs: the number of principal components. Bold values indicates the optimal prediction results for each mechanical parameter.
