Partly Interpretable Machine Learning Method of Ginseng Geographical Origins Recognition and Analysis by Hyperspectral Measurements

Meng LI; Xiao-bo ZHANG; Shao-bo LIU; Xing-feng CHEN; Lu-qi HUANG; Ting-ting SHI; Rui YANG; Shu LIU; Feng-jie ZHENG

doi:10.3964/j.issn.1000-0593(2022)04-1217-05

Journals >Spectroscopy and Spectral Analysis >Volume 42 >Issue 4 >Page 1217 > Article

Spectroscopy and Spectral Analysis
Vol. 42, Issue 4, 1217 (2022)

Partly Interpretable Machine Learning Method of Ginseng Geographical Origins Recognition and Analysis by Hyperspectral Measurements

Meng LI^{1、1; 2;}, Xiao-bo ZHANG^2、2;, Shao-bo LIU^3、3;, Xing-feng CHEN^{4、4; *;}, Lu-qi HUANG^{5、5; *;}, Ting-ting SHI^2、2;, Rui YANG^6、6;, Shu LIU^7、7;, and Feng-jie ZHENG^8、8;

Author Affiliations

¹1. School of Pharmacy, Henan University of Chinese Medicine, Zhengzhou 450046, China

²2. State Key Laboratory Breeding Base of Dao-di Herbs, National Resource Center for Chinese Materia Medica, Chinese Academy of Chinese Medical Sciences, Beijing 100700, China

³3. Big Date Center, Space Star Technology Co., Ltd., Beijing 100086, China

⁴4. State Environmental Protection Key Laboratory of Satellite Remote Sensing, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

⁵5. State Key Laboratory Breeding Base of Dao-di Herbs, Chinese Academy of Chinese Medical Sciences, Beijing 100700, China

⁶6. Key Laboratory of Remote Sensing of Gansu Province, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou 730000, China

⁷7. Jilin Provincial Key Laboratory of Chinese Medicine Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China

⁸8. School of Space Information, Space Engineering University, Beijing 101416, China

show less

DOI: 10.3964/j.issn.1000-0593(2022)04-1217-05 Cite this Article

Meng LI, Xiao-bo ZHANG, Shao-bo LIU, Xing-feng CHEN, Lu-qi HUANG, Ting-ting SHI, Rui YANG, Shu LIU, Feng-jie ZHENG. Partly Interpretable Machine Learning Method of Ginseng Geographical Origins Recognition and Analysis by Hyperspectral Measurements[J]. Spectroscopy and Spectral Analysis, 2022, 42(4): 1217 Copy Citation Text

show less

Example of ginseng hyperspectral imagery(a): True color composite image acquired by VNIR-1024 camera; (b): Color composite image acquired by SWIR-384 camera

Fig. 1. Example of ginseng hyperspectral imagery
(a): True color composite image acquired by VNIR-1024 camera; (b): Color composite image acquired by SWIR-384 camera

Download full size | View in the Article

Fig. 2. Spectral reflectance curves from different geographical origins

Download full size | View in the Article

产地划分	运行次数										平均总体精度
产地划分	1	2	3	4	5	6	7	8	9	10	平均总体精度
东北与否	100	91	100	100	100	91	100	100	100	100	98.2
四省	100	60	80	80	80	80	100	80	80	80	82
八地	80	40	80	60	80	100	60	80	20	80	68

Table 1. The recognition accuracies under three origin classification scale (100%)

View in the Article

产地划分	波段范围/nm	重要程度/%	备注
东北与否	1 000~2 500 7001 000	89 7	最重要波段为1 798 nm, 重要性为6.5%, 其他无显著重要波段。比红光波长短的波段无用
四省	1 0002 500 4801 000	81 15	无显著重要波段, 重要性最大值2.4%, 紫、蓝光无用
八地	4001 000 1 0001 750	43.7 52.3	无显著重要波段, 重要性最大值1.4%, 训练样本少, 结论可参考价值低

Table 2. The feature bands statistics of ginseng origin recognition by random forest

Download Citation

Save the article for my favorites

Paper Information