• Spectroscopy and Spectral Analysis
  • Vol. 40, Issue 11, 3567 (2020)
Ye HUANG1、1, Li LIU1、1, Jing LIANG1、1, Hong-xia YANG1、1, Xiao-li LI1、1, and Ning XU1、1
Author Affiliations
  • 1[in Chinese]
  • 11. Department of Pharmacy, Zhejiang Skin Disease Prevention and Treatment Center, Deqing 313200, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2020)11-3567-06 Cite this Article
    Ye HUANG, Li LIU, Jing LIANG, Hong-xia YANG, Xiao-li LI, Ning XU. Research on the Visualization Differentiation of Atractylodes Lancea Granule Manufactures Based on Hyperspectral Imaging Technology Combined With the Selection of Characteristic Wavelengths[J]. Spectroscopy and Spectral Analysis, 2020, 40(11): 3567 Copy Citation Text show less
    (a)The TLC result and (b) average hyperspectral of Atractylodes Lancea granulesSample 1, 2, and 3 were manufactured from Huarun Sanjiu, Jiangyin Tianjiang, and Zhejiang Huisong respectively
    Fig. 1. (a)The TLC result and (b) average hyperspectral of Atractylodes Lancea granules
    Sample 1, 2, and 3 were manufactured from Huarun Sanjiu, Jiangyin Tianjiang, and Zhejiang Huisong respectively
    (a) The changing trend of the number of sampled variables, (b) RMSECV values, (c) regression coefficients of each variable with the increasing of sampling runs of CARS, (d) results of selected characteristic wavelengths by RF, (e) discriminant rate and (f) kappa coefficient in different models based on different wavelengths
    Fig. 2. (a) The changing trend of the number of sampled variables, (b) RMSECV values, (c) regression coefficients of each variable with the increasing of sampling runs of CARS, (d) results of selected characteristic wavelengths by RF, (e) discriminant rate and (f) kappa coefficient in different models based on different wavelengths
    Characteristic wavelength selection results based on (a) preliminary selection and (b) CA method and (c) correlation analysis among characteristic wavelengths selected by CARS in the distinguish study of Atractylodes Lancea granules from different manufactures
    Fig. 3. Characteristic wavelength selection results based on (a) preliminary selection and (b) CA method and (c) correlation analysis among characteristic wavelengths selected by CARS in the distinguish study of Atractylodes Lancea granules from different manufactures
    The distinguish result map of Atractylodes Lancea granules from different manufacturers based on CARS-CA-LS-SVM model
    Fig. 4. The distinguish result map of Atractylodes Lancea granules from different manufacturers based on CARS-CA-LS-SVM model
    波长选择方法数量特征波长/nm
    CARS19975, 1 122, 1 220, 1 237, 1 294, 1 368, 1 372, 1 402, 1 409, 1 415, 1 419,
    1 439, 1 442, 1 446, 1 476, 1 540,
    1 561, 1 611, 1 669
    CARS-CA4975, 1 220, 1 419, 1 476
    RF51 005, 1 368, 1 409, 1 415, 1 442
    RF-CA21 005, 1 442
    SFS10924, 975, 1 005, 1 294, 1 348, 1 365,
    1 419, 1 442, 1 584, 1 645
    SFS-CA4924, 1 005, 1 419, 1 584
    SPA8941, 948, 954, 1 126, 1 146, 1 274,
    1 412, 1 483
    SPA-CA3948, 1 146, 1 412
    Table 1. Selected characteristic wavelengths in the distinguish study of Atractylodes Lancea granules from different manufactures based on hyperspectral technology
    波长选
    择方法
    波长
    数目
    KNNBPNNPLS-DALS-SVM
    总体判
    别率/%
    Kappa
    系数
    总体判
    别率/%
    Kappa
    系数
    总体叛
    别率/%
    Kappa
    系数
    γδ2总体判
    别率/%
    Kappa
    系数
    CARS19900.84100110011.904×1020.289 31001
    CARS-CA4940.91980.9710011.198×1030.2891001
    RF5880.68100110014.083×1030.025 051001
    RF-CA2860.78780.67820.724.08×1030.025 0860.79
    SFS10900.84100110013.664×1051.4121001
    SFS-CA4900.84980.97680.503.66×1051.411001
    SPA8940.91980.9710016.889×1030.509 01001
    SPA-CA3940.91640.42860.786.89×1030.509940.91
    Table 2. Model discrimination based on characteristic wavelengths in the distinguish study of Atractylodes Lancea granules from different manufactures
    Ye HUANG, Li LIU, Jing LIANG, Hong-xia YANG, Xiao-li LI, Ning XU. Research on the Visualization Differentiation of Atractylodes Lancea Granule Manufactures Based on Hyperspectral Imaging Technology Combined With the Selection of Characteristic Wavelengths[J]. Spectroscopy and Spectral Analysis, 2020, 40(11): 3567
    Download Citation