基于高光谱成像技术的宁夏枸杞产地溯源鉴别

摘要/Abstract

摘要： 宁夏枸杞因其特有的感官品质和药用价值，市场上假冒宁夏枸杞产品现象频发。因此，快速鉴别枸杞产地对生产者、消费者和市场经济至关重要。本研究目的是基于高光谱成像（400~1000 nm）结合化学计量学开发一种用于识别枸杞产地多元化的检测方法。获取宁夏、甘肃、内蒙、青海和新疆5个不同产地的枸杞高光谱图像，并基于阈值分割方法从感兴趣区域提取光谱数据。同时使用多种预处理方法来消除光谱的干扰信息，研究表明基于归一化反射光谱（Normalized reflectance spectrum, NR）的判别模型表现出较好的性能。进一步地采用连续投影算法（Succesive projections algorithm, SPA）、竞争性自适应重加权算法（Competitive adaptive reweighted sampling, CARS）、粒子群优化算法（Particle swarm optimization, PSO）、迭代保留信息变量算法（Iteratively retaining informative variables, IRIV）和CARS+IRIV选择特征波长。研究结果表明CARS+IRIV选取波长建立的简化模型性能最优，从二元分类到五元分类模型，特征波长仅占全波长的15.6%~27.7%，预测集准确率分别为97.7%、90.9%、89.2%、87.1%。此外，为了更加直观辨别分类种类，使用混淆矩阵可视化最佳简化分类模型。在对宁夏枸杞分类中获得了令人满意的灵敏度、特异性和Kappa系数。结果表明，高光谱成像技术结合化学计量学方法可有效鉴别枸杞产地，可为枸杞产业发展提供关键技术支撑。

关键词: 高光谱成像, 枸杞, 产地鉴别, 特征波长

Abstract: Ningxia Lycium barbarum have unique organoleptic qualities and medicinal properties, and counterfeit Ningxia Lycium barbarum products are pervasive in the market. As a result, it is vital for producers, consumers and market economy to quickly identify the geographical origin of Lycium barbarum. This study aimed to develop a detection method based on hyperspectral imaging (400~1000 nm) combined with chemometrics to identify the diverse geographical origins of Lycium barbarum. Hyperspectral images of Lycium barbarum from five different production areas in Ningxia, Gansu, Inner Mongolia, Qinghai and Xinjiang were acquired, and spectral data was extracted from the region of interest based on threshold segmentation method. Multiple preprocessing methods were employed to eliminate the interference information from the spectra, and the performance showed that the discriminative model based on normalized reflectance spectrum (NR) exhibited better performance. Furthermore, the successive projections algorithm (SPA), competitive adaptive reweighted sampling (CARS), particle swarm optimization (PSO), iteratively retaining informative variables (IRIV), and CARS+IRIV were used to select characteristic wavelengths. The results showed that the simplified model, which utilized CARS+IRIV for wavelength selection, achieved the best performance. In models ranging from binary to quintuple classifications, the selected characteristic wavelengths accounted for only 15.6% to 27.7% of the full spectra. The correct classification rates obtained in the prediction set were 97.7%, 90.9%, 89.2%, and 87.1%, respectively. In addition, a confusion matrix was employed to visualize the optimal simplified classification model and intuitively distinguish the classification categories. Satisfactory sensitivity, specificity and Kappa coefficients were obtained in classifying Ningxia Lycium barbarum. The results illustrated that hyperspectral imaging technology combined with chemometric methods can effectively identify the geographical origin of Lycium barbarum and provide crucial technical support for the development of Lycium barbarum industry.

Key words: Hyperspectral imaging, Lycium Barbarum, Geographical origin identification, characteristic wavelengths

中图分类号:

袁伟东姜洪喆杨诗雨张聪周禹周宏平. 基于高光谱成像技术的宁夏枸杞产地溯源鉴别[J]. 食品科学, 0, (): 0-0.

weidong yuan Hong-Zhe JIANG. Geographical Origin Identification of Ningxia Lycium Barbarum Using Hyperspectral Imaging Technology[J]. FOOD SCIENCE, 0, (): 0-0.

参考文献

[1]徐常青, 刘赛, 徐荣, 等.我国枸杞主产区生产现状调研及建议[J].中国中药杂志, 2014, 39(11):1979-1984
[2]卢有媛, 郭盛, 张芳, 等.枸杞属药用植物资源系统利用与产业化开发[J].中国现代中药, 2019, 21(1):29-36
[3]加帕尔如克亚, 孙玉敬, 钟烈州, 等.枸杞植物化学成分及其生物活性的研究进展[J].中国食品学报, 2013, 13(8):161-172
[4]李艳萍, 刘冰, 刘楠, 等.不同种质宁夏枸杞红外光谱特征及其与表型性状的相关性分析[J].食品研究与开发, 2023, 44(9):178-182
[5]ZHAO L Q, QIU Z Q, NARASIMHAMOORTHY B, et al.Development of a rapid, high-throughput method for quantification of zeaxanthin in Chinese wolfberry using HPLC–DAD[J]. Industrial Crops and Products, 2013, 47: 51-57. DOI:10.1016/j.indcrop.2013.02.008.
[6]MENG J, LIU Z, GOU C L, et al.Geographical origin of Chinese wolfberry (goji) determined by carbon isotope analysis of specific volatile compounds[J]. Journal of Chromatography B, 2019, 1105: 104-112. DOI:10.1016/j.jchromb.2018.12.011.
[7]LI Q, YU X, XU L, et al.Novel method for the producing area identification of Zhongning Goji berries by electronic nose[J]. Food Chemistry, 2017, 221: 1113-1119. DOI:10.1016/j.foodchem.2016.11.049.
[8]CHENG L juan, LIU G shan, HE J guo, et al.Development of a novel quantitative function between spectral value and metmyoglobin content in Tan mutton[J]. Food Chemistry, 2021, 342: 128351. DOI:10.1016/j.foodchem.2020.128351.
[9]YUAN W, JIANG H, SUN M, et al.Geographical Origin Identification of Chinese Tomatoes Using Long-Wave Fourier-Transform Near-Infrared Spectroscopy Combined with Deep Learning Methods[J]. Food Analytical Methods, 2023. https://doi.org/10.1007/s12161-023-02444-1. DOI:10.1007/s12161-023-02444-1.
[10]BACA-BOCANEGRA B, NOGALES-BUENO J, HERNáNDEZ-HIERRO J M, et al.Evaluation of extractable polyphenols released to wine from cooperage byproduct by near infrared hyperspectral imaging[J]. Food Chemistry, 2018, 244: 206-212. DOI:10.1016/j.foodchem.2017.10.027.
[11]JIANG H, YUAN W, RU Y, et al.Feasibility of identifying the authenticity of fresh and cooked mutton kebabs using visible and near-infrared hyperspectral imaging[J]. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy, 2022, 282: 121689. DOI:10.1016/j.saa.2022.121689.
[12]HE J, CHEN L, CHU B, et al.Determination of Total Polysaccharides and Total Flavonoids in Chrysanthemum morifolium Using Near-Infrared Hyperspectral Imaging and Multivariate Analysis[J].Molecules, 2018, 23(9):2395-
[13]王磊, 覃鸿, 李静, 等.近红外高光谱图像的宁夏枸杞产地鉴别[J].光谱学与光谱分析, 2020, 40(4):1270-1275
[14]孙俊, 金夏明, 毛罕平, 等.高光谱图像技术在掺假大米检测中的应用[J].农业工程学报, 2014, 30(21):301-307
[15]周宏平, 胡逸磊, 姜洪喆, 等.基于高光谱成像的油茶籽含油率检测方法[J].农业机械学报, 2021, 52(5):308-315
[16]ARAúJO M C U, SALDANHA T C B, GALV?O R K H, et al.The successive projections algorithm for variable selection in spectroscopic multicomponent analysis[J].Chemometrics and Intelligent Laboratory Systems, 2001, 57(2):65-73
[17]LI H, LIANG Y, XU Q, et al.Key wavelengths screening using competitive adaptive reweighted sampling method for multivariate calibration[J].Analytica Chimica Acta, 2009, 648(1):77-84
[18]WANG S, SUN J, FU L, et al.Identification of red jujube varieties based on hyperspectral imaging technology combined with CARS-IRIV and SSA-SVM[J].Journal of Food Process Engineering, 2022, 45(10):e14137-
[19]梁琨, 刘全祥, 潘磊庆, 等.基于高光谱和-算法的‘库尔勒香梨’可溶性固形物含量检测[J].南京农业大学学报, 2018, 41(4):760-766
[20]GALLO M.Discriminant partial least squares analysis on compositional data[J].Statistical Modelling, 2010, 10(1):41-56
[21]TANG N, SUN J, YAO K, et al.Identification of Lycium barbarum varieties based on hyperspectral imaging technique and competitive adaptive reweighted sampling-whale optimization algorithm-support vector machine[J].Journal of Food Process Engineering, 2021, 44(1):e13603-
[22]FATCHURRAHMAN D, NOSRATI M, AMODIO M L, et al.Comparison Performance of Visible-NIR and Near-Infrared Hyperspectral Imaging for Prediction of Nutritional Quality of Goji Berry (Lycium barbarum L)[J].Foods, 2021, 10(7):1676-
[23]JI Y, SUN L, LI Y, et al.Non-destructive classification of defective potatoes based on hyperspectral imaging and support vector machine[J]. Infrared Physics & Technology, 2019, 99: 71-79. DOI:10.1016/j.infrared.2019.04.007.
[24]姜洪喆, 杨雪松, 李兴鹏, 等.油茶果自然霉变程度的可见近红外与中短波近红外光谱检测[J].食品科学, 2023, 44(4):272-277
[25]QIAO M, XU Y, XIA G, et al.Determination of hardness for maize kernels based on hyperspectral imaging[J]. Food Chemistry, 2022, 366: 130559. DOI:10.1016/j.foodchem.2021.130559.
[26]徐新刚, 赵春江, 王纪华, 等.基于可见光-近红外新光谱特征和最优组合原理的大麦叶片氮含量监测[J].红外与毫米波学报, 2013, 32(4):351-358
[27]ZHANG C, JIANG H, LIU F, et al.Application of Near-Infrared Hyperspectral Imaging with Variable Selection Methods to Determine and Visualize Caffeine Content of Coffee Beans[J].Food and Bioprocess Technology, 2017, 10(1):213-221