Chinese Optics, Volume. 17, Issue 1, 128(2024)
Optimal position for suger content detection of Yongquan honey oranges based on hyperspectral imaging technology
Fig. 3. Spectral curves of Yongquan honey orange. (a) Original spectral curves of different parts; (b) average spectral curves of different parts
Fig. 4. Selecting process of the characteristic wavelength of the fruit stem part by CARS. (a) Changes in number of variables; (b) changes in the RMSECV; (c) changes in regression coefficient
Fig. 5. Location map of the characteristic wavelengths in the fruit stem part based on the CARS algorithm corresponding to the pretreatments (a) Baseline and (b) MSC
Fig. 6. Location maps of the characteristic wavelengths based on CARS algorithm. (a) Calyx; (b) equator and (c) global
Fig. 7. Stability values of the fruit stem part after UVE screening
Fig. 8. Location maps of characteristic wavelengths in the fruit stem part based on the UVE algorithm corresponding to (a) Baseline and (b) MSC
Fig. 9. Location maps of the characteristic wavelengths of the UVE-based algorithm. (a) Calyx; (b) equator and (c) global
Fig. 10. Scatter plots of the Yongquan honey oranges sugar content prediction models (a) MSC-CARS-PLSR and (b) MSC-CARS-LSSVM
Table 1. Statistical analysis of the sugar content of different parts of Yongquan honey orange
View tableView in ArticleTable 1. Statistical analysis of the sugar content of different parts of Yongquan honey orange
蜜桔部位 样本数 最大值 / OBrix 最小值 / OBrix 平均值 / OBrix 标准差 / OBrix 花萼 120 19.8 10.8 15.2 1.39 果茎 120 17.9 10.1 14.2 1.52 赤道 120 18.2 11.3 14.5 1.37 全局 120 17.8 11.2 14.6 1.34
Table 2. Comparison of PLSR models for detecting the sugar content of Yongquan honey orange based on different pretreatments
View tableView in ArticleTable 2. Comparison of PLSR models for detecting the sugar content of Yongquan honey orange based on different pretreatments
预测模型 预处理方法 建模集 预测集 RC RMSEC/OBrix RP RMSEP/OBrix 花萼模型 Raw 0.946 0.384 0.893 0.457 SNV 0.847 0.58 0.806 0.688 MSC 0.832 0.622 0.766 0.564 Baseline 0.921 0.409 0.890 0.518 SG 0.932 0.427 0.898 0.436 果茎模型 Raw 0.949 0.428 0.859 0.587 SNV 0.902 0.593 0.882 0.669 MSC 0.889 0.599 0.864 0.587 Baseline 0.931 0.498 0.913 0.468 SG 0.943 0.455 0.868 0.569 赤道模型 Raw 0.932 0.471 0.861 0.553 SNV 0.946 0.408 0.936 0.370 MSC 0.960 0.365 0.878 0.458 Baseline 0.964 0.349 0.933 0.384 SG 0.924 0.497 0.861 0.555 全局模型 Raw 0.971 0.305 0.920 0.388 SNV 0.945 0.403 0.901 0.435 MSC 0.953 0.374 0.934 0.435 Baseline 0.926 0.469 0.855 0.495 SG 0.927 0.476 0.923 0.384
Table 3. Comparison of LSSVM models for detecting the sugar content of Yongquan honey orange basedon different pretreatments
View tableView in ArticleTable 3. Comparison of LSSVM models for detecting the sugar content of Yongquan honey orange basedon different pretreatments
预测模型 预处理 方法 建模集 预测集 RC RMSEC/OBrix RP RMSEP/ OBrix 花萼模型 Raw 0.921 0.470 0.860 0.513 SNV 0.938 0.383 0.789 0.700 MSC 0.959 0.323 0.788 0.539 Baseline 0.942 0.360 0.869 0.585 SG 0.923 0.459 0.876 0.477 果茎模型 Raw 0.979 0.286 0.782 0.750 SNV 0.908 0.594 0.834 0.710 MSC 0.955 0.404 0.884 0.596 Baseline 0.924 0.527 0.642 0.854 SG 0.953 0.419 0.827 0.650 赤道模型 Raw 0.965 0.355 0.829 0.594 SNV 0.954 0.388 0.906 0.405 MSC 0.973 0.315 0.827 0.530 Baseline 0.979 0.281 0.867 0.544 SG 0.956 0.388 0.845 0.575 全局模型 Raw 0.962 0.355 0.892 0.443 SNV 0.972 0.296 0.897 0.456 MSC 0.980 0.253 0.946 0.400 Baseline 0.973 0.293 0.811 0.590 SG 0.961 0.356 0.909 0.414
Table 4. Comparison of PLSR and LSSVM models for different parts of honey oranges after CARS characteristic wavelengths screening
View tableView in ArticleTable 4. Comparison of PLSR and LSSVM models for different parts of honey oranges after CARS characteristic wavelengths screening
预测模型 不同部位 建模集 预测集 RC RMSEC/ OBrix RP RMSEP/ OBrix PLSR 花萼 0.926 0.447 0.918 0.400 果茎 0.928 0.507 0.922 0.424 赤道 0.933 0.452 0.914 0.400 全局 0.948 0.394 0.942 0.399 LSSVM 花萼 0.927 0.445 0.914 0.408 果茎 0.951 0.412 0.904 0.546 赤道 0.960 0.352 0.901 0.423 全局 0.975 0.274 0.955 0.395
Table 5. Comparison of PLSR and LSSVM models for different parts of honey oranges after UVE characteristic wavelengths screening
View tableView in ArticleTable 5. Comparison of PLSR and LSSVM models for different parts of honey oranges after UVE characteristic wavelengths screening
预测模型 不同部位 建模集 预测集 RC RMSEC/ OBrix RP RMSEP/ OBrix PLSR 花萼 0.890 0.538 0.850 0.519 果茎 0.885 0.633 0.812 0.655 赤道 0.943 0.419 0.933 0.364 全局 0.949 0.393 0.937 0.434 LSSVM 花萼 0.901 0.514 0.838 0.537 果茎 0.950 0.416 0.896 0.575 赤道 0.950 0.400 0.900 0.423 全局 0.956 0.368 0.943 0.414
Get Citation
Copy Citation Text
Bin LI, Xia WAN, Ai-lun LIU, Ji-ping ZOU, Ying-jun LU, Chi YAO, Yan-de LIU. Optimal position for suger content detection of Yongquan honey oranges based on hyperspectral imaging technology[J]. Chinese Optics, 2024, 17(1): 128
Category:
Received: Mar. 30, 2023
Accepted: --
Published Online: Mar. 28, 2024
The Author Email:
