发光学报, 2019, 40 (8): 1055, 网络出版: 2019-09-02   

可见/近红外高光谱成像技术对长枣中葡萄糖含量的无损检测

Non-destructive Detective of Glucose Contect in Lingwu Jujube by Vis/NIR Hyperspectral Imaging Technology
作者单位
宁夏大学 农学院, 宁夏 银川 750021
摘要
利用可见/近红外高光谱成像技术结合化学计量学方法建立长枣中葡萄糖含量的预测模型, 为灵武长枣中的葡萄糖含量快速无损检测提供了一种科学方法。采用可见/近红外(400~1 000 nm)高光谱采集灵武长枣的光谱数据, 利用HPLC测量长枣中的葡萄糖含量; 样本经过剔除异常值、样本集划分后对原始光谱采用6种预处理; 对优选出的最佳预处理方法使用7种方法降维处理, 建立全波段和特征波长的PLSR、MLR预测模型。结果表明, SG(7)为最佳预处理方法, Rc=0.826 5, Rp=0.791 0; 利用CARS、IRF、SPA、BiPLS、UVE、IRF+CARS、BiPLS+CARS分别选出18, 61, 7, 51, 15, 33, 27个特征波长; PLSR-IRF+CARS模型最优, Rc=0.835 3, Rp=0.832 2。实验结果证明高光谱成像技术对长枣中的葡萄糖含量的快速无损检测是可行的。
Abstract
The Vis/NIR hyperspectral imaging technique combined with chemometrics method was applied to build a glucose contect prediction model, which provided a scientific method for the rapid non-destructive detection of glucose content in Lingwu long jujube. The spectral data were acquired by Vis/NIR hyperspectral imaging(400-1 000 nm). Simultaneously, high-performance liquid chromatography was used to detect the glucose content; after eliminating abnormal sample data and dividing into sample set, the 6 pretreatment methods were used to process original spectrum. Then the characteristic wavelength was selected from the pretreatmented spectral data by 7 methods, and the PLSR and MLR prediction models were established based on full wavelength and feature wavelength, respectively. The results show that the SG(7)method is the optimal pretreatment method; and the number of the characteristic wavelengths by CARS, IRF, SPA, BiPLS, UVE, IRF+CARS and BiPLS+CARS are 18, 61, 7, 51, 15, 33, 27, respectively; the IRF+CARS model was the best among the models developed, and its Rc=0.835 3, Rp=0.832 2. So it is feasible to predicte glouse content of Lingwu long jujube based on hyperspectral imaging technique.

程丽娟, 刘贵珊, 万国玲, 何建国. 可见/近红外高光谱成像技术对长枣中葡萄糖含量的无损检测[J]. 发光学报, 2019, 40(8): 1055. CHENG Li-juan, LIU Gui-shan, WAN Guo-ling, HE Jian-guo. Non-destructive Detective of Glucose Contect in Lingwu Jujube by Vis/NIR Hyperspectral Imaging Technology[J]. Chinese Journal of Luminescence, 2019, 40(8): 1055.

本文已被 6 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!