近红外光谱技术在农产品品质检测中的应用

近红外光谱技术是一种新型的无损检测技术,在许多领域都得到了很好的应用。本文从农产品中各种物质成分含量预测、分类鉴别、腐烂鉴别、实时监测几个方面综述了近红外光谱技术在农产品品质检测上的应用,并对其在仪器硬件的研究和开发、化学计量学方法的探索与研究以及快速在线检测方法的研究等方面的发展趋势进行了展望。     

Detection of low-level gluten content in flour and batter by near infrared reflectance spectroscopy (NIRS)

Near infrared spectroscopy (NIRS) has been used as a valuable tool for quality control in the food industry. The aim of the present study was to investigate the possibility of developing a NIRS calibration for gluten determination in flour and batter, suitable for the analysis of gluten-free food products. Reflectance data was used for calibration based on modified partial least squares (MPLS) regression. Independent prediction equations were developed for flour and for batter. Spectral models using mean spectra of two scans (average spectra), were compared with those using the two individual spectral data. The best model obtained for flour was using the average spectral data (R² = 0.985; r² = 0.967) and for batter samples was using the individual spectral data (R² = 0.926; r² = 0.825). It is concluded that the application of NIRS methodology can predict accurately the concentration of gluten content in flours and batters, but it should not be considered as a reliable method for determining gluten contamination in gluten-free products.     

Integration of independent component analysis with near infrared spectroscopy for evaluation of rice freshness

The storage time and conditions of rice has an enormous effect on its appearance, flavor, and quality of the nutrients; and the acidity of rice usually increases with prolonged storage. Therefore, evaluation of freshness is an important issue for rice quality. In this study, the NIR (near infrared) spectra combined with independent component analysis (ICA) technique was used to evaluate the rice freshness. A total of 180 white rice samples were collected from 6 crop seasons for the purpose of developing an ICA-NIR based procedure for rice freshness as quantified by pH values. Values of pH were determined by a BTB-MR (bromothymol blue – methyl red) method. The best calibration model of white rice was developed using the smoothed first derivative spectra, five ICs and cross-validation; the results indicated that r² (coefficient of determination) = 0.924, and in units of pH, SEC (standard error of calibration) = 0.145, SEP (standard error of prediction) = 0.146, bias = 0.001, and RPD (residual predictive deviation) = 3.65. Freshness of white rice could be distinguished either visually by a 3-dimensional diagram composed from ICs 2, 3 and 4, or statistically by a calibration model. The results show that ICA with NIR has the potential to be adopted as an effective method for evaluating rice freshness.     

Non-destructive prediction of chemical composition in sunflower seeds by near infrared spectroscopy

Near infrared reflectance spectroscopy (NIRS) was explored as a technique to predict moisture (M), oil and crude protein (CP) content on intact sunflower seeds (Helianthus annuus L.). Three hundred samples were scanned intact in a monochromator instrument NIRS 6500 (NIRSystems, Silver Spring, MD, USA). Calibration equations were developed using modified partial least square regression (MPLS) with internal cross validation. Samples were split in two sets, one set used as calibration (n=250) where the remaining samples (n=50) were used as validation set. Two mathematical treatments (first and second derivative), none (log 1/R) and standard normal variate and detrend (SNVD) as scatter corrections were explored. The coefficient of determination in calibration (R_cal²) and the standard error in cross validation (SECV) were 0.95 (SECV: 3.3) for M; 0.96 (SECV: 13.1) for CP and 0.90 (SECV: 22.3) for oil in g kg⁻¹ on a dry weight basis (second derivative, 400–2500 nm). Prediction models accounted for less than 65, 70 and 72% of the total variation for oil, M and CP, respectively. However, it was concluded that NIRS is a suitable technique to be used as a tool for rapid pre-screening of quality characteristics on breeding programs.     

A chemometric evaluation of the underlying physical and chemical patterns that support near infrared spectroscopy of barley seeds as a tool for explorative classification of endosperm genes and gene combinations

Near infrared spectroscopic (NIR; 1100–2500 nm), chemical and genetic data were combined to study the pleiotropic secondary effects of mutant genes on milled samples in a barley seed model. NIR and chemical data were both effective in classifying gene and gene combinations by Principal Component Analysis (PCA). Risø mutants R-13, R-29 high (1→3, 1→4)-β-glucan, low starch and R-1508 (high lysine, reduced starch), near isogeneic controls and normal lines and recombinants were studied. Based on proteome analysis results, six anti-microbial proteins were followed during endosperm development revealing pleiotropic gene effects in expression timing that supporting the gene classification. To verify that NIR spectroscopy data represents a physio–chemical fingerprint of the barley seed, physical and chemical spectral components were partially separated by Multiple Scatter Correction and their genetic classification ability verified. Wavelength bands with known water binding and (1→3, 1→4)-β-glucan assignments were successfully predicted by partial least squares regression giving insight into how NIR-data works in classification. Highly reproducible gene-specific, covariate, pleiotropic classification patterns from NIR and chemical data were demonstrated in PCAs and by visual inspection of NIR spectra. Thus PCA classification of NIR-data gives the classical genetic concept, ‘pleiotropy’, a new operational definition as a fingerprint from a spectroscopic representation of the phenome carrying genetic, physical and chemical information. It is concluded that barley seed phenotyping by NIR and chemometrics is a new, reliable tool for characterising the pleiotropic effects of mutant gene combinations and other genotypes in selecting barley for quality in plant breeding.     

Amino acid profile of the quinoa (Chenopodium quinoa Willd.) using near infrared spectroscopy and chemometric techniques

The high content of amino acids of the quinoa, especially essential amino acids (higher than other cereals) makes a food increasingly demanded by consumers. A total of twelve amino acids (arginine, cystine, isoleucine, leucine, lysine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine and valine) were analyzed in quinoa samples from Chile by near infrared spectroscopy (NIR) with direct application to the samples of a remote fiber-optic reflectance probe. The calibration results using modified partial least squares (MPLS) regression satisfactorily allowed the determination of the concentrations of this amino acid group with high multiple correlation coefficients (RSQ = 0.97–0.71) and low standard prediction errors (SEPC = 0.07–0.20). The prediction capacity (RPD) for the arginine, the cystine, the isoleucine, the lysine, the serine, the threonine, the tryptophan, the tyrosine and the valine ranged between 2.6 and 5.2, for the rest of amino acids were higher to 1.8, indicating that the NIRS equations obtained were applicable to unknown samples. It has confirmed that NIRS technology is a method that may be useful to replace the traditional methods for routine analysis of some amino acids.     

利用近红外技术预测向日葵籽仁品质性状

为实现向日葵育种材料的品质性状快速预测,选取154份向日葵籽仁样品,结合化学测定值和近红外光谱,利用化学计量学手段建立向日葵籽仁品质指标的近红外模型,评估其在向日葵籽仁粗蛋白、粗脂肪、油酸、亚油酸等重要品质性状测定中的可行性。结果表明,改进偏最小二乘法建模效果最佳,其粗脂肪、粗蛋白、油酸、亚油酸、饱和脂肪酸及不饱和脂肪酸的定标相关系数分别为0.975、0.950、0.973、0.951和0.913,交叉验证相关系数分别 为0.969、0.939、0.915、0.927和0.711。用检验集对模型进行验证,粗脂肪、蛋白质、油酸、亚油酸、饱和及不饱和脂肪酸的外部检验相关系数（R2）分别为0.959、0.950、0.937、0.906和0.930。本研究建立的模型质量较高,能够满足向日葵籽仁品质成分的快速测定,可为向日葵品质育种前期大量、快速的筛选提供技术支持。     

一种国产近红外仪分析油菜籽三种主要品质参数研究

利用中国农业科学院油料作物研究所研制的NYDL-3000智能型多参数粮油品质速测仪分析了油菜籽含油量、硫甙和芥酸含量,测定结果与传统化学方法进行了比较,其中含油量结果与PC-120核磁共振仪和MATRIX-I傅立叶变换近红外光谱仪做验证。内部交叉验证结果显示：含油量、硫甙和芥酸含量的决定系数R2分别为0.954 3、0.985 3和0.982 3,均方差分别为0.0469 、0.419 8和0.160 4,T检验显示三种品质参数交叉验证集的近红外光谱法（NYDL-3000）预测值和化学值差异均不显著。外部检验结果表明：硫甙含量、含油量两种方法测定值的相关系数分别为0.837 6和0.951 7,平均绝对误差分别为7.54％和0.02％,均在平均绝对误差允许的范围之内;成对数据双尾t检验表明,近红外光谱法和化学法两者结果差异不显著。芥酸分析中,NYDL-3000将342份样品中3份低芥酸材料判定为高芥酸,近红外光谱法和化学法两者结果符合率达99.12%。含油量的测定结果显示：NYDL-3000和PC-120、MATRIX-I结果的相关系数分别为0.956 6和0.901 5,平均绝对误差分别为0.78％和3.06％。研究结果表明,利用NYDL-3000智能型多参数粮油品质速测仪所建立的品质模型可以满足油菜育种对早代材料芥酸、硫甙含量和含油量的快速测定要求。     

花生籽仁含油量近红外模型的构建及其应用

花生(Arachis hypogaea L.)籽仁含油量是花生品质评价的重要指标,建立快速高效的含油量检测方法,对加快高油花生品种选育意义重大。本研究选用高油亲本宇花14(含油量59.32%)与低油亲本LOP215(含油量48.97%)杂交构建的RIL群体为建模材料,使用Thermo公司(美国)生产的Antaris Ⅱ型傅立叶变换近红外光谱分析仪对229份样品籽仁进行光谱采集,随后测定籽仁含油量。利用偏最小二乘法(partial least squares,PLS)构建花生籽仁含油量近红外定标模型,该模型的内部验证均方差(root mean square error of cross validation, RMSECV)为0.885,相关系数R²=0.9147。选用未参与建模的21份花生材料对该模型进行外部验证,模型预测值和化学测定值的决定系数R²=0.9492,表明该模型可适用于花生籽仁含油量检测。利用该模型对宇花14与LOP215杂交后代群体进行筛选,获得含油量超过55%的优良株系21个,含油量低于48%的株系9个,可为花生高低含油量品...     

Comparison of near infrared reflectance spectroscopy calibration strategies for the botanical composition of grass‐clover mixtures

Predicting species composition in mixed swards by near infrared reflectance spectroscopy (NIRS) can save labour in grassland research, provided equations are available. This study compares calibration strategies to predict species composition in swards with tall fescue (Festuca arundinacea), perennial ryegrass (Lolium perenne) and white clover (Trifolium repens). The compared calibration strategies were based on either real or artificial samples. Real samples are samples taken from multispecies swards; the species composition is known by hand separating; after separating, the original samples are recomposed. Artificial samples are samples obtained by mixing single species grown in pure stands in known proportions. The performance of the equation based on real samples was significantly better than the performance of the equation based on artificial samples. We hypothesized that the weak performance of the equation based on artificial samples was due to a lack of environmental variation in the spectra of the artificial samples. The hypothesis was supported by the good performance of a novel calibration strategy based on the spectra of the artificial samples with added variation. According to the obtained results, a calibration strategy based on few but diverse calibration samples is discussed.     

Biorefining of perennial grasses: Development of visible and near infrared tools for assessing fineness of extracted fibre

The biorefining of grass offers an opportunity to integrate primary production agriculture with the extraction of fibre. The study was aimed at developing protocols for processing ryegrass to determine fibre content and to generate visible and near infrared reflectance (Vis-NIR) calibrations for estimating fibre fineness with the aid of a new reference airflow method. The method for determining fineness of processed fibres has been adapted from flax fibre protocols. A Vis-NIR calibration using partial least squares (PLS) regression method was employed to generate models with a calibration set consisting of 85 samples obtained from fresh and ensiled grasses. The PLS model was successfully validated with 21 independent samples with a prediction error of 1.26 dtex. An optimised PLS model (r² = 0.86) consisting of 106 samples has been developed. The quality assurance protocols could be used for assessing fibre content and quality of silage, hay and fresh grass.     

Feasibility analysis of NIR for detecting sweet corn seeds vigor

This paper explores the feasibility of particle-based detection and grading of seed vigor based on a self-built seed single-granulation device using near infrared spectroscopy (NIRS). Sweet corn with uniform kernel size was used for this study. The seed samples were divided into three types, they were normal seeds, artificially aged seeds and heat-damaged seeds. A 2-part spectral acquisition of each seed were performed, one for the collection of seeds that fall into the detection zone within the separation pipe, another was on the static platform, whose collection was performed on 5 faces of each seed. Partial least squares discriminant analysis (PLS-DA) was used to classify the original data of the seeds. In the 2 parts, the discriminant results of the unprocessed normal seeds and the artificial accelerated aging seeds, the untreated normal seeds and the heat-damaged seeds showed that classification accuracy was higher than 98%. The research indicates that the spectral data of different positions of seeds can reflect their activity information, and it is feasible to detect and classify seeds in real time in the detection area of the separation pipeline.     

Visible-near infrared spectroscopy based citrus greening detection: Evaluation of spectral feature extraction techniques

Citrus greening is a serious disease affecting citrus production in Florida and different parts of the world. This disease is spread by an insect vector and the trees are killed several years after infection. There is no known treatment for the disease. Disease detection and removal of infected trees is a critical part of citrus greening disease management efforts. This paper reports the evaluation of spectral features extracted from visible-near infrared spectroradiometer spectra for their potential to detect citrus greening disease. The extraction of spectral features is an effort to lower the cost of the optical sensor while maintaining their performance. Spectral features: (i) spectral reflectance bands and (ii) vegetation indices (VIs) were derived from 350-2,500 nm spectral reflectance data using two feature extraction methods: stepwise discriminant analysis and stepwise regression analysis. Following the selection of spectral features, the features were assessed using two classifiers, quadratic discriminant analysis (QDA) and soft independent modeling of classification analogies (SIMCA) to determine the overall and individual class classification accuracies. The classification results indicated that both the spectral features (spectral bands and VIs) yielded good overall (higher than 80%) and healthy class (higher than 85%) classification accuracies using the QDA-based algorithm. The SIMCA-based algorithm yielded good average citrus greening class classification accuracy (higher than 83%) using selected spectral features. Thus, the present study demonstrates the applicability of utilizing spectral features for detection of greening in citrus.     

高油玉米群体油分、蛋白质和淀粉含量近红外分析模型的构建

以BHO高油玉米F2∶3家系为材料,应用主成分空间和傅里叶变换近红外光谱分析技术,采用偏最小二乘回归法(PLS),建立了测定高油玉米子粒的油分、蛋白质和淀粉含量的近红外校正模型。预处理分别采用一阶导数 矢量归一化、一阶导数 多元散射校正及直线相减等方法,主成分维数分别为5、9、9。验证分析表明,所建立的油分、蛋白质和淀粉含量的校正模型的校正和预测效果最好,其校正决定系数(R2cal)分别为0.950、0.973、0.976,交叉验证决定系数(R2cv)和外部验证决定系数(R2val)在0.918～0.948,各项误差(RMSEE、RMSECV、RMSEP)在0.305%～0.721%。结果表明,所建立的高油玉米完整子粒品质性状三成分模型的准确度和精确度均较高,可以满足高油玉米群体大量样品无损品质分析的需要。     

Impact of sampling and storage technique,and duration of storage,on the composition of fresh grass when analysed using near‐infrared reflectance spectroscopy

The objective of this study was to examine the effect of sampling technique (pluck or cut), storage duration (immediate analysis, 24‐h or 48‐h), storage temperature (ambient or chilled) and storage conditions (air present, air excluded or breathable) on the composition of fresh grass sampled from a sward managed to simulate grazing. Treatments were repeated across four sampling dates, with grass samples stored in grip seal bags prior to analysis using near‐infrared reflectance spectroscopy. Grass sampled by ‘pluck’ had a higher crude protein and ME content, and a lower acid detergent fibre (ADF) content, compared to that sampled by ‘cut’. Grass stored for 48 h had a lower water soluble carbohydrate (WSC) and ME content and a higher ADF content than for immediate analysis. Samples stored for 24 h did not differ from immediate analysis. Grass stored at ambient temperature had a lower WSC and ME content compared to immediate analysis. Grass stored under ‘breathable’ conditions had a lower ME content and higher ADF content than immediate analysis or samples stored with air present or air excluded. It is recommended that grass for analysis should be sampled by cutting, stored chilled (4°C) in a sealed bag to minimize exposure to oxygen and analysed within 24 h of harvest.     

Application of two-dimensional cross-correlation spectroscopy to analyse infrared (MIR and NIR) spectra recorded during bread dough mixing

During dough mixing chemical, biochemical and physical transformations occur that allow dough formation to be characterized by common chemical and biochemical methods. Recently, spectrometric methods were used to characterize the dough mixing. The Mid-infrared (MIR) and the Near-infrared (NIR) spectroscopy allow information concerning chemical content and composition of food products to be obtained. The aim of this study is to apply FT-NIR and FT-MIR spectroscopy to monitor dough chemical changes, and to correlate those signals by the 2D Cross-Correlation (2D CORR) method. The 2D CORR was used to emphasize chemical assignment of the NIR band modifications (particularly for protein) during dough mixing.The 2D CORR analysis of the raw NIR and MIR spectra demonstrated that five NIR regions are highly correlated to protein vibrations. The 2D CORR analysis of the NIR and MIR spectra after second derivative demonstrated that the amide bands present high R² for the NIR bands at (1189–1216), (1351–1474) and (1873) nm. A low R² is obtained between the amide I and amide II bands and the (2026–2123) and (2280–2325) nm regions. The amide III band presents a slightly higher R² for those NIR regions.The 2D CORR analysis of NIR and MIR spectra allow more specific NIR regions associated to chemical modifications of protein structure to be identified. The 2D CORR analysis of the second derivative spectra is more precise for the identification of the NIR regions implied in dough mixing compared to the 2D CORR analysis of raw NIR and MIR spectra.     

Estimation of dry matter and starch concentration in potatoes by determination of under-water weight and near infrared spectroscopy

Summary Potato food and non-food industries need information about dry matter (DM) and starch concentration of tubers. Therefore, calculation of dry matter concentration and starch concentration via under-water weight were optimised. Determination coefficients (R²) were 0.92 and 0.83 (starch concentration between 13 and 23%), and 0.94 and 0.88 (starch concentration ≥13%) for dry matter and starch concentration, respectively. In a second attempt, near infrared spectroscopy models for both constituents were calculated (R² of validation set was 0.98 and 0.96 for dry matter and starch concentration, respectively). Results pointed out superiority of the near infrared technique with less divergence than the techniques of under-water weighting.     

Agronomic evaluation of high-protein and high-oil soybean genotypes for specialty markets

Soybeans (Glycine max (L.) Merr.) are mainly grown for protein and oil purposes. The objectives of this study were to: 1) evaluate maturity group (MG) IV and V soybean genotypes for traits associated with local adaptability (yield, plant height, and maturity), and 2) determine seed protein and oil concentrations, their correlation with yield, and respective heritabilities. A total of 40 MG IV or V genotypes were evaluated in four Arkansas locations in 2008 and 2009. The results showed that the genotype x year x location effect was significant for all traits studied, except maturity. Protein and oil concentrations were negatively correlated (?0.91) and highly heritable (0.89–0.93 and 0.82–0.83, respectively). Four promising high-yielding genotypes with moderately high to high protein or oil levels were identified: R05-4682 (high protein) and R05-4256 (high oil) in the MG IV test, and R05-1772 (high protein) and R05-71 (high oil) in the MG V test. These genotypes could potentially bring added value to Arkansas farmers’ fields.