食品与发酵工业 >

2023 , Vol. 49 >Issue 14: 295 - 301

DOI: https://doi.org/10.13995/j.cnki.11-1802/ts.033379

分析与检测

基于融合预处理的掺假咖啡太赫兹光谱识别

  • 张傲林 ,
  • 李绅 ,
  • 邹颖芳 ,
  • 王继芬 ,
  • 张震 ,
  • 孙一健
展开
  • 1(中国人民公安大学 侦查学院,北京,100038)
    2(武汉体育学院 运动训练学院,湖北 武汉,430079)
本科生(王继芬教授为通信作者,E-mail:wangjifen58@126.com)

收稿日期: 2022-08-22

  修回日期: 2022-09-08

  网络出版日期: 2023-08-30

收起

Terahertz spectral identification of adulterated coffee based on fusion pretreatment

  • ZHANG Aolin ,
  • LI Shen ,
  • ZOU Yingfang ,
  • WANG Jifen ,
  • ZHANG Zhen ,
  • SUN Yijian
Expand
  • 1(People's Public Security University of China, College of Investigation, Beijing 100038, China)
    2(Sports Training College of Wuhan Institute of Physical Education, Wuhan Sports University, Wuhan 430079, China)

Received date: 2022-08-22

  Revised date: 2022-09-08

  Online published: 2023-08-30

Fold

摘要

该研究通过太赫兹时域光谱采集70组含有西布曲明成分的咖啡在0~2.5 THz频段的光谱信息,建立随机森林、支持向量机、贝叶斯判别分析3种模式识别方法并进行比较研究。结果表明,未经过预处理的模型识别准确率较低。选择一阶导数、二阶导数、不同类型的巴特沃斯滤波器和Pearson特征选择融合光谱方法进行光谱信号处理。基于一阶导数处理的贝叶斯判别分析模型准确率为98.6%,基于高通巴特沃斯滤波器的随机森林模型分类准确率为94.2%,基于特征提取的融合光谱支持向量机(support vector machine,SVM)模型分类准确率为100%。选择最优预处理的SVM模型进一步对同一品牌不同地区的掺假咖啡进行鉴别,准确率为100%。研究实现了“品牌-产地”的二级特征识别,可为公安机关打击涉及咖啡的食品安全犯罪提供参考。

关键词： 咖啡; 西布曲明; 太赫兹时域光谱; 模式识别; 光谱预处理

本文引用格式

张傲林 , 李绅 , 邹颖芳 , 王继芬 , 张震 , 孙一健 . 基于融合预处理的掺假咖啡太赫兹光谱识别[J]. 食品与发酵工业, 2023 , 49(14) : 295 -301 . DOI: 10.13995/j.cnki.11-1802/ts.033379

Abstract

The spectral information of 70 groups of coffee containing sibutramine in 0-2.5 THz band was collected by terahertz time-domain spectroscopy. Three pattern recognition methods, including random forest, support vector machine, and Bayesian discriminant analysis, were established and compared. The model recognition accuracy without preprocessing was lower. The first derivative, second derivative, different types of Butterworth filters, and Pearson feature selection fusion spectroscopy method were selected for spectral signal processing. The accuracy of the Bayesian discriminant analysis model based on first derivative processing was 98.6%, the classification accuracy of the random forest model based on the high-pass Butterworth filter was 94.2%, and the classification accuracy of the fusion spectral support vector machine (SVM) model based on feature extraction was 100%. The SVM model with optimal pretreatment was selected to further identify adulterated coffee in different areas of the same brand, and the accuracy rate was 100%. The two-level feature recognition of “brand-origin” was realized, which could provide certain help for the public security organs to crack down on food safety crimes involving coffee.

Key words： coffee; sibutramine; THz time-domain spectroscopy; pattern recognition; spectra pretreatment

参考文献

[1] CONNOLEY I P, LIU Y L, FROST I, et al.Thermogenic effects of sibutramine and its metabolites[J].British Journal of Pharmacology, 1999, 126(6):1 487-1 495.
[2] ARTERBURN D E, CRANE P K, VEENSTRA D L.The efficacy and safety of sibutramine for weight loss[J].Archives of Internal Medicine, 2004, 164(9):994.
[3] JAMES W, CATERSON I D, COUTINHO W, et al.Effect of sibutramine on cardiovascular outcomes in overweight and obese subjects[J].The New England Journal of Medicine, 2010, 363(10):905-917.
[4] 邵金良, 杨东顺, 樊建麟, 等.高效液相色谱—紫外双波长同时测定咖啡及咖啡制品中10种化合物含量[J].食品科学, 2016, 37(12):128-133.
SHAO J L, YANG D S, FAN J L, et al.Simultaneous determination of 10 polyphenolic and alkaloidal components in coffee and coffee-based products by HPLC-double wavelength UV detection[J].Food Science, 2016, 37(12):128-133.
[5] 陈华舟, 许丽莉, 秦强.光谱分析法应用于咖啡成分的测定和咖啡种类的鉴别[J].光谱学与光谱分析, 2014(6):1 688-1 692.
CHEN H Z, XU L L, QIN Q.Spectroscopic methods applied to component determination and species identification for coffee[J].Spectroscopy and Spectral Analysis, 2014(6):1 688-1 692.
[6] 陈秀明, 奚星林, 潘丙珍, 等.基于近红外光谱技术的咖啡掺假快速鉴别方法[J].现代食品科技, 2018, 34(10):253-257;290.
CHEN X M, XI X L, PAN B Z, et al.Study of the rapid identification method of coffee adulteration based on near-infrared spectroscopy[J].Modern Food Science & Technology, 2018, 34(10):253-257;290.
[7] 王艳艳, 何勇, 邵咏妮, 等.基于可见-近红外光谱的咖啡品牌鉴别研究[J].光谱学与光谱分析, 2007, 27(4):702-706.
WANG Y Y, HE Y, SHAO Y N, et al.Discrimination among different brands of coffee by using vis-near infrared spectra[J].Spectroscopy and Spectral Analysis, 2007, 27(4):702-706.
[8] DRUZ·BICKI K, MIELCAREK J, KIWILSZA A, et al.Computationally assisted (solid-state density functional theory) structural (X-ray) and vibrational spectroscopy (FT-IR, FT-RS, TDs-THz) characterization of the cardiovascular drug Lacidipine[J].Crystal Growth & Design, 2015, 15(6):2 817-2 830.
[9] ZHAN H L, WU S X, BAO R M, et al.Qualitative identification of crude oils from different oil fields using terahertz time-domain spectroscopy[J].Fuel, 2015, 143:189-193.
[10] 张红涛, 蔡永胜, 王宇, 等.太赫兹时域光谱在农业领域的应用研究进展[J/OL].激光与光电子学进展, 2022.http://kns.cnki.net/kcms/detail/31.1690.tn.20220308.0950.004.html.
ZHANG H T,CAI Y S,WANG Y,et al.Research progress on the application of terahertz time domain spectroscopy in agriculture[J/OL].Laser & Optoelectronics Progress, 2022.http://kns.cnki.net/kcms/detail/31.1690.tn.20220308.0950.004.html.
[11] 杨少壮, 李灿, 李辰, 等.基于太赫兹光谱分析技术的肉类鉴别[J].食品与发酵工业, 2021, 47(11):227-235.
YANG S Z, LI C, LI C, et al.Study of meat identification based on terahertz spectroscopy[J].Food and Fermentation Industries, 2021, 47(11):227-235.
[12] 王倩, 葛宏义, 蒋玉英, 等.基于太赫兹时域光谱的大米品种识别研究[J].食品工业科技, 2022,43(2):19-25.
WANG Q, GE H Y, JIANG Y Y, et al.Identification of rice varieties based on terahertz time domain spectroscopy[J].Science and Technology of Food Industry, 2022,43(2):19-25.
[13] 李天莹. 太赫兹光谱技术在食品添加安全领域的应用与研究[D].南京:南京林业大学, 2020.
LI T Y.Application and research of terahertz spectroscopy in the field of food additive safety[D].Nanjing:Nanjing Forestry University, 2020.
[14] 张仲雄, 张东莉, 田世杰, 等.太赫兹波谱技术在食品掺假检测中的研究进展[J].光谱学与光谱分析, 2021, 41(5):1 379-1 386.
ZHANG Z X, ZHANG D L, TIAN S J, et al.Research progress of terahertz spectroscopy technique in food adulteration detection[J].Spectroscopy and Spectral Analysis, 2021, 41(5):1 379-1 386.
[15] 曾明飞, 朱玉杰, 冯国红, 等.基于可见/近红外光谱的蓝莓新鲜度快速评价研究[J].食品与发酵工业, 2022,48(20):252-259.
ZENG M F, ZHU Y J, FENG G H, et al.Rapid evaluation of blueberry freshness based on visible/near-infrared spectroscopy[J].Food and Fermentation Industries, 2022,48(20):252-259.
[16] 古锟山, 王继芬, 曾啸虎.基于滤波器-光谱数据降维的指甲地区识别[J].分析测试学报, 2022, 41(5):746-753;760.
GU K S, WANG J F, ZENG X H.Recognition of fingernail region based on filter-spectral feature extraction[J].Journal of Instrumental Analysis, 2022, 41(5):746-753;760.
[17] 张震, 王继芬, 刘津彤.基于先进光谱融合技术-特征优化的复印纸无损识别[J/OL].分析测试学报, 2021.http://kns.cnki.net/kcms/detail/44.1318.TH.20210707.0858.002.html.
ZHANG Z, WANG J F, LIU J T.Non-destructive recognition of copy paper based on advanced spectral fusion and feature optimization[J/OL].Journal of Instrumental Analysis, 2021.http://kns.cnki.net/kcms/detail/44.1318.TH.20210707.0858.002.html.
[18] 王梓笛, 李双妹, 李艳, 等.基于拉曼光谱-支持向量机的乳制品快速智能鉴别技术研究[J].光谱学与光谱分析, 2020, 40(S1):157-158.
WANG Z D, LI S M, LI Y, et al.Rapid and intelligent identification of dairy products based on Raman spectroscopy and support vector machine[J].Spectroscopy and Spectral Analysis, 2020, 40(S1):157-158.
[19] 何欣龙, 王继芬, 何亚, 等.Bayes判别的塑钢窗红外光谱快速识别[J].激光杂志, 2019, 40(11):33-37.
HE X L,WANG J F,HE Y, et al.Infrared spectroscopy identification of plastic steel windows based on Bayes discrimination analysis[J].Laser Journal, 2019, 40(11):33-37.
[20] 何欣龙, 王继芬.牛顿插值多项式-导数光谱无损检测车用保险杠[J].激光技术, 2020, 44(3):333-337.
HE X L, WANG J F.The identification about the automotive bumper based on Newton interpolation polynomial-infrared derivative spectroscopy[J].Laser Technology, 2020, 44(3):333-337.
[21] 蒋强. 基于THz-TDS的聚乙烯管道厚度测量与缺陷检测技术研究[D].赣州:江西理工大学, 2019.
JIANG Q.Research on thickness measurement and defect detection technology of polyethylene pipeline based on THz-TDS[D].Ganzhou:Jiangxi University of Science and Technology, 2019.
[22] 闫政旭, 秦超, 宋刚.基于Pearson特征选择的随机森林模型股票价格预测[J].计算机工程与应用, 2021, 57(15):286-296.
YAN Z X, QIN C, SONG G.Random forest model stock price prediction based on Pearson feature selection[J].Computer Engineering and Applications, 2021, 57(15):286-296.
[23] 邱薇纶, 周燕舞, 石孟良.基于数据融合策略植物油光谱模式的识别[J/OL].中国油脂, 2022.http://kns.cnki.net/kcms/detail/61.1099.TS.20220510.1008.004.html.
QIU W L, ZHOU Y W, SHI M L.Spectral pattern recognition of the vegetable oils based on data fusion strategy[J/OL].China Oils and Fats, 2022.http://kns.cnki.net/kcms/detail/61.1099.TS.20220510.1008.004.html.
Options
文章导航

/

〈 〉

技术支持：北京玛格泰克科技发展有限公司