Wood Species Recognition Based on Visible and Near-Infrared Spectral Analysis Using Fuzzy Reasoning and Decision-Level Fusion
Table 11
Overall recognition accuracy (ORA) and time requirement (TR) comparisons of the testing dataset for different classifiers (ORA and TR refer to the mean values of 20 tests).
Model
Dataset
ORA (%)
TR (ms)
BC
VIS (PCA7dim)
76.20
0.050256
NIR (PCA7dim)
77.80
0.037920
VIS (TSNE7dim)
55.00
0.047946
NIR (TSNE7dim)
61.60
0.052136
RF
VIS (PCA7dim)
83.80
2.437714
NIR (PCA7dim)
84.00
2.326404
VIS (TSNE7dim)
76.40
2.654658
NIR (TSNE7dim)
80.40
2.310316
BPN
VIS (PCA7dim)
63.84
0.135030
NIR (PCA7dim)
61.48
0.117026
VIS (TSNE7dim)
54.04
0.123924
NIR (TSNE7dim)
60.80
0.140712
LIBSVM
VIS (PCA7dim)
83.60 (linear)
0.092936
54.80 (RBF)
0.123525
NIR (PCA7dim)
84.40 (linear)
0.081890
80.20 (RBF)
0.131751
VIS (TSNE7dim)
72.40 (linear)
0.086184
23.80 (RBF)
0.253162
NIR (TSNE7dim)
77.60 (linear)
0.091526
23.00 (RBF)
0.211325
LeNet-5
VIS
92.40
0.763721
FRC
VIS (PCA4dim)
90.20
0.137274
NIR (PCA4dim)
92.92
0.141672
VIS (TSNE4dim)
71.88
0.159326
NIR (TSNE4dim)
81.12
0.173528
D-S FRC
VIS + NIR (PCA4dim)
93.84
3.026516
FW-D-S FRC
VIS + NIR (PCA4dim)
94.76
3.071578
VIS: visible band; NIR: near-infrared band; FRC: fuzzy reasoning classifier; PCA: principal component analysis; BC: Bayes classifier; RF: random forest; BPN: BP network; CNN: convolutional neural network; LIBSVM: support vector machine; T-SNE: T-distributed stochastic neighbor embedding; D-S: Dempster–Shafer.
