Non-Destructive Detection and Visualization of Soybean Moisture Content Using Hyperspectral Technique

Fig. 3. Selection process of CARS variables
(a): Variation trend of the number of variables with the number of samples; (b): RMSECV; (c): The change process of regression coefficient of each variable with sampling times (The blue line represents the position with the lowest RMSECV)

Download full size

View in Article

Fig. 4. Stability distribution curve of UVE-PLSR modle

Download full size

View in Article

Fig. 5. Visualization of soybean moisture content
(a): Huadou 2; (b): Kendou 40; (c): Wandou 701; (d): Wandou 34

Download full size

View in Article

Table 1. Moisture content of soybean samples
View table
View in Article
Table 1. Moisture content of soybean samples
样本集样本数/个水分含量/%
最大值最小值平均值标准偏差
校正集 72 11.06 6.12 7.86 1.63
预测集 24 10.69 6.13 7.99 1.36
总样本 96 11.06 6.12 7.90 1.58

Table 2. PLSR model based on different pretreatment methods

View table

View in Article

Table 2. PLSR model based on different pretreatment methods

预处理方法	PCs	校正集		交互验证集
预处理方法	PCs	$R_{C}^{2}$	RMSEC	$R_{CV}^{2}$	RMSECV
无	8	0.957 6	0.278	0.926 6	0.373
Moving Average	8	0.956 7	0.281	0.924 3	0.379
S-G平滑	8	0.957 3	0.279	0.925 9	0.378
Baseline	8	0.958 7	0.275	0.930 4	0.369
Normalize	8	0.960 7	0.268	0.938 0	0.353
SNV	9	0.961 1	0.266	0.921 1	0.388
MSC	7	0.949 4	0.304	0.916 8	0.384
Detrending	8	0.962 2	0.263	0.930 3	0.364

Table 3. Performance of models based on different pretreatment methods and characteristic wavelengths selecting methods

View table

View in Article

Table 3. Performance of models based on different pretreatment methods and characteristic wavelengths selecting methods

No	模型	波长数	校正集		交互验证集		预测集
No	模型	波长数	$R_{C}^{2}$	RMSEC	$R_{CV}^{2}$	RMSECV	$R_{P}^{2}$	RMSEP
1	PLSR	216	0.957 6	0.278	0.926 6	0.373	0.957 1	0.329
2	PCR	216	0.953 7	0.291	0.930 0	0.367	0.963 7	0.303
3	SVMR	216	0.955 6	0.287	0.911 8	0.402	0.886 2	0.537
4	SPA-PLSR	14	0.967 4	0.244	0.933 7	0.358	0.972 9	0.262
5	SPA-PCR	14	0.967 7	0.243	0.934 1	0.355	0.972 9	0.262
6	SPA-SVMR	14	0.955 8	0.287	0.927 0	0.367	0.906 1	0.488
7	CARS-PLSR	16	0.982 9	0.177	0.968 8	0.254	0.952 0	0.349
8	CARS-PCR	16	0.982 5	0.179	0.964 4	0.257	0.955 8	0.335
9	CARS-SVMR	16	0.953 7	0.294	0.931 5	0.356	0.915 5	0.463
10	UVE-PLSR	29	0.964 7	0.254	0.944 0	0.350	0.953 8	0.299
11	UVE-PCR	29	0.967 3	0.244	0.944 0	0.326	0.958 5	0.324
12	UVE-SVMR	29	0.936 8	0.340	0.903 8	0.420	0.915 5	0.463
13	Normalize-SPA-PLSR	14	0.974 3	0.217	0.948 3	0.325	0.977 8	0.238
14	Normalize-SPA-PCR	14	0.974 6	0.215	0.948 9	0.313	0.977 8	0.238

Tools

Get Citation

Copy Citation Text

Cheng-qian JIN, Zhen GUO, Jing ZHANG, Cheng-ye MA, Xiao-han TANG, Nan ZHAO, Xiang YIN. Non-Destructive Detection and Visualization of Soybean Moisture Content Using Hyperspectral Technique[J]. Spectroscopy and Spectral Analysis, 2022, 42(10): 3052

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites