Author Affiliations
11. College of Mechanical and Electrical Engineering, Tarim University, Alar 843300, China22. College of Plant Science, Tarim University, Alar 843300, Chinashow less
Fig. 1. Layout of test samples
Fig. 2. Background spectra of some samples
Fig. 3. Eliminating abnormal spectra of winter Jujube by Mahalanobis distance
Fig. 4. Eliminating abnormal sugar content samples of winter Jujube by concentration residual
Fig. 5. CARS method selects the characteristic wavelength of winter Jujube sugar
(a): Relationship between interation number and wavelength variable; (b): RMSECV value for different interations; (c): Variable PLS regression coefficient value
Fig. 6. Background spectra of winter jujube and results of con+1st modeling of sugar content
(a): Model sample regression; (b): Model sample error
Fig. 7. Comparison of inversion results of Roujean model
(a): Winter jujube; (b): Red grapes; (c): Fragrant pear
Fig. 8. Comparison of inversion results of Walthall model
(a): Winter jujube; (b): Red grapes; (c): Fragrant pear
Fig. 9. Error comparison of different fruits retrieved by Roujean model
(a): Winter jujube; (b): Red grapes; (c): Fragrant pear
Fig. 10. Error comparison of different fruits retrieved by Walthall model
(a): Winter jujube; (b): Red grapes; (c): Fragrant pear
处理方法 | r | rmsec | rmsep |
---|
Con | 0.549 93 | 2.40 | 2.72 | MSC | 0.814 82 | 1.67 | 2.43 | SNV | 0.827 29 | 1.62 | 2.34 | Con+1st | 0.853 31 | 1.50 | 2.29 | MSC+1st | 0.806 89 | 1.70 | 2.26 | SNV+1st | 0.803 86 | 1.71 | 2.28 | Con+2st | 0.723 36 | 2.00 | 2.26 | MSC+2st | 0.751 78 | 1.92 | 2.16 | SNV+2st | 0.746 10 | 1.91 | 2.42 |
|
Table 1. Results of background spectrum and sugar modeling of winter Jujube under different treatment methods
理化值 | 预处理方法 | r | rmsec | rmsep |
---|
糖度 | Con+1st | 0.853 31 | 1.50 | 2.29 | SNV | 0.827 29 | 1.62 | 2.34 | MSC | 0.814 82 | 1.67 | 2.43 | 水分 | Con | 0.741 28 | 0.551 | 0.950 | SNV | 0.669 93 | 0.610 | 1.10 | MSC | 0.664 99 | 0.614 | 1.11 | 探测角 | Con+2st | 0.985 58 | 0.227 | 0.404 | Con | 0.975 40 | 0.296 | 0.260 | Con+1st | 0.974 72 | 0.300 | 0.323 | 方位角 | Con+2st | 0.941 83 | 2.84 | 5.72 | MSC+1st | 0.925 33 | 3.21 | 4.33 | SNV+1st | 0.914 64 | 3.42 | 4.35 | 相位角 | MSC+2st | 0.960 94 | 0.810 | 1.38 | SNV+2st | 0.960 42 | 0.804 | 1.48 | Con+2st | 0.957 55 | 0.860 | 1.45 |
|
Table 2. Modeling results of different winter jujube physicochemical values
理化值 | 预处理方法 | r | rmsec | rmsep |
---|
糖度 | Con+2st | 0.822 67 | 0.460 | 0.645 | MSC+1st | 0.818 21 | 0.465 | 1.05 | SNV+1st | 0.816 46 | 0.467 | 1.02 | 水分 | Con+1st | 0.784 74 | 0.577 | 0.955 | Con+2st | 0.721 79 | 0.644 | 1.09 | MSC | 0.616 77 | 0.733 | 1.11 | 探测角 | Con+1st | 0.992 73 | 0.463 | 0.633 | Con | 0.992 41 | 0.473 | 0.633 | Con+2st | 0.990 57 | 0.527 | 0.803 | 方位角 | Con | 0.910 45 | 2.50 | 3.99 | MSC+1st | 0.894 78 | 2.70 | 5.24 | SNV+1st | 0.890 70 | 2.75 | 4.70 | 相位角 | Con | 0.957 01 | 1.20 | 1.20 | Con+1st | 0.955 67 | 1.22 | 1.32 | SNV+1st | 0.950 29 | 1.29 | 1.77 |
|
Table 3. Modeling results of different Red grapes physicochemical values
理化值 | 预处理方法 | r | rmsec | rmsep |
---|
糖度 | Con+2st | 0.913 34 | 0.505 | 1.02 | MSC+2st | 0.909 23 | 0.524 | 0.982 | SNV+2st | 0.904 02 | 0.530 | 0.965 | 水分 | SNV+2st | 0.891 33 | 0.412 | 0.885 | MSC+1st | 0.868 04 | 0.451 | 1.02 | SNV | 0.720 12 | 0.631 | 1.09 | 探测角 | Con | 0.974 69 | 0.588 | 0.692 | Con+2st | 0.966 59 | 0.674 | 0.661 | SNV | 0.964 62 | 0.694 | 0.666 | 方位角 | MSC | 0.936 88 | 3.76 | 5.55 | SNV | 0.934 54 | 3.82 | 5.56 | MSC+1st | 0.900 14 | 4.69 | 6.30 | 相位角 | Con | 0.956 27 | 1.25 | 1.91 | SNV | 0.950 56 | 1.33 | 1.70 | MSC | 0.946 25 | 1.38 | 1.71 |
|
Table 4. Modeling results of different Fragrant Pear physicochemical values
样品 编号 | 冬枣 | 红提 | 香梨 |
---|
R2 | r | rmsep | 误差/% | R2 | r | rmsep | 误差/% | R2 | r | rmsep | 误差/% |
---|
1 | 0.968 8 | 0.992 6 | 0.028 0 | 4.41 | 0.975 0 | 0.989 4 | 0.031 6 | 9.12 | 0.981 5 | 0.982 1 | 0.041 2 | 2.92 | 2 | 0.932 8 | 0.989 7 | 0.042 8 | 8.84 | 0.952 6 | 0.984 1 | 0.038 8 | 9.27 | 0.756 2 | 0.986 5 | 0.076 0 | 12.65 | 3 | 0.925 0 | 0.989 6 | 0.044 4 | 7.20 | 0.897 8 | 0.988 4 | 0.058 2 | 14.32 | 0.876 5 | 0.957 4 | 0.046 9 | 4.70 | 4 | 0.853 2 | 0.979 5 | 0.065 2 | 9.89 | 0.834 0 | 0.977 0 | 0.073 7 | 16.60 | 0.746 2 | 0.967 1 | 0.075 0 | 11.03 | 5 | 0.958 4 | 0.996 7 | 0.033 8 | 7.54 | 0.918 8 | 0.966 1 | 0.062 0 | 8.36 | 0.814 8 | 0.962 1 | 0.069 4 | 10.57 | 6 | 0.989 8 | 0.996 6 | 0.017 1 | 2.80 | 0.975 9 | 0.988 7 | 0.029 2 | 7.03 | 0.892 9 | 0.964 5 | 0.046 9 | 4.65 | 7 | 0.971 6 | 0.991 7 | 0.027 6 | 3.92 | 0.900 0 | 0.991 2 | 0.062 9 | 14.31 | 0.814 9 | 0.963 7 | 0.066 7 | 9.53 | 8 | 0.875 6 | 0.985 3 | 0.059 7 | 13.55 | 0.971 0 | 0.985 9 | 0.033 2 | 9.13 | 0.762 1 | 0.969 6 | 0.079 6 | 12.88 |
|
Table 5. Comparison of inversion results of roujean model
样品 编号 | 冬枣 | 红提 | 香梨 |
---|
R2 | r | rmsep | 误差/% | R2 | r | rmsep | 误差/% | R2 | r | rmsep | 误差/% |
---|
1 | 0.977 6 | 0.992 5 | 0.024 0 | 2.62 | 0.909 2 | 0.977 3 | 0.056 5 | 13.89 | 0.767 7 | 0.975 3 | 0.079 7 | 7.62 | 2 | 0.912 7 | 0.995 4 | 0.047 3 | 8.92 | 0.883 1 | 0.961 5 | 0.064 2 | 10.94 | 0.912 0 | 0.988 4 | 0.050 6 | 3.73 | 3 | 0.931 5 | 0.998 6 | 0.042 9 | 8.51 | 0.685 9 | 0.963 6 | 0.105 2 | 29.63 | 0.849 4 | 0.985 0 | 0.067 4 | 7.89 | 4 | 0.964 6 | 0.997 5 | 0.030 9 | 5.83 | 0.796 0 | 0.957 6 | 0.084 7 | 17.19 | 0.790 7 | 0.970 8 | 0.081 3 | 9.99 | 5 | 0.873 7 | 0.971 5 | 0.058 4 | 8.51 | 0.752 7 | 0.949 7 | 0.094 8 | 20.09 | 0.889 0 | 0.955 6 | 0.059 0 | 6.76 | 6 | 0.985 8 | 0.992 9 | 0.019 6 | 3.38 | 0.964 6 | 0.988 1 | 0.035 9 | 8.47 | 0.913 0 | 0.975 1 | 0.052 1 | 6.00 | 7 | 0.979 0 | 0.996 0 | 0.023 8 | 4.25 | 0.911 5 | 0.988 2 | 0.056 7 | 14.41 | 0.820 2 | 0.988 0 | 0.074 9 | 10.67 | 8 | 0.921 6 | 0.990 2 | 0.045 9 | 7.52 | 0.974 6 | 0.988 3 | 0.030 4 | 8.59 | 0.770 3 | 0.923 0 | 0.084 6 | 10.07 |
|
Table 6. Comparison of inversion results of Walthall model