基于傅里叶近红外光谱和电子鼻技术的苹果霉心病无损检测

为研究傅里叶近红外光谱技术(Fourier transform near infrared spectroscopy,FT-NIRS)和电子鼻技术分别结合化学计量学方法对苹果霉心病的判别效果,以“红富士”霉心病苹果和健康苹果为试材,利用近红外光谱技术,基于主成分分析建立Fisher判别和多层感知器(multi-layer perceptron,MLP)神经网络模型;同时利用电子鼻技术分别结合Fisher判别、MLP神经网络和径向基函数神经网络3种化学计量学的方法建立判别模型。根据建模集和验证集的预测准确率综合考虑,基于主成分分析建立的MLP神经网络模型和电子鼻结合MLP神经网络模型对苹果霉心病的判别效果最好,验证集中的正确判别率分别达到87.7%和86.2%。说明电子鼻和近红外光谱技术均可以较好地判别苹果霉心病。     

苹果品种及损伤苹果的FT-NIR鉴别研究

用傅里叶近红外光谱技术(FT-NIR)对不同品种的苹果以及损伤嘎啦和完好嘎啦进行快速、无损检测,比较不同判别方法对所建立的区分苹果品种及苹果损伤模型的影响。结果表明:损伤嘎啦和完好嘎啦的近红外图谱经小波分析预处理后,用12000～4000cm-1波数范围的前5个主成分分别结合多层感知神经网络、径向基神经网络、Fisher判别3种方法所建立的判别模型对未知样本的正确判别率分别为97.8%、87.2%和84.8%,基于权重法用多元线性回归(MLR)所选择的特征波长所建立的Fisher判别模型对未知样本的正确判别率为89.1%;用偏最小二乘判别(PLS-DA)所建立的判别模型对未知样本的正确判别率为100%,由于PLS-DA模型对训练集和验证集的正确判别率均为100%,因此PLS-DA模型优于其他模型。不同品种苹果的光谱经平滑预处理后,用全波数范围12000～4000cm-1的前6个主成分所建立的判别模型优于经验波数范围8000～4500cm-1所建立的判别模型,其较优模型对建模集和验证集的正确判别率分别为90.9%和92.1%。近红外光谱技术结合化学计量学可以快速、无损鉴别苹果是否有损伤以及不同品种的苹果。     

基于透射光谱的苹果霉心病多因子无损检测

针对目前苹果霉心病难以检测的问题,提出一种基于透射光谱的苹果霉心病多因子无损检测方法,通过融合多波段透射光谱与苹果直径,构建苹果霉心病判别模型,实现了苹果霉心病无损检测。搭建光谱测试范围在200～1 025 nm的透射光谱采集平台,实验获取232 个苹果样本的透射光谱数据,采用游标卡尺获得苹果直径数据。采用杂散光校正,非线性校正对苹果透射原始光谱进行预处理,选取与霉心病发病相关的12 个波段透射光强值,结合苹果的直径进行主成分分析,将分析的结果作为自变量,建立苹果霉心病Fisher判别模型。经过异校验验证,模型总体识别率为93.1%,而仅采用透射光谱构建的模型识别率为91.37%。结果表明,基于透射光谱与直径结合的多因子检测方法可实现苹果霉心病的准确判定,为苹果霉心病的快速、无损检测提供可行思路。     

基于高光谱反透射图像的新疆冰糖心红富士水心鉴别

采用高光谱成像技术结合化学计量法,采集新疆冰糖心红富士好果与水心病果样本在波长范围380～1 004 nm的可见近红外高光谱反透射图像,选取感兴趣区域获得平均光谱,对原始光谱采用直接差分一阶求导等9种光谱预处理方法,再分别用主成分分析、快速独立分量分析、相关系数法完成数据降维,结合贝叶斯判别、K最近邻法、马氏距离判别、最小二乘支持向量机、二次线性判别方法识别是否有水心病。结果表明,主成分分析提取前15主成分,采用标准正态变量变换-主成分分析-最小二乘支持向量机与多元散射校正-主成分分析-最小二乘支持向量机模型识别效果最优,校正集和预测集识别率分别为100%和91.2%。     

利用可见/近红外光谱技术对小米产地进行溯源研究

目的利用可见/近红外光谱技术对产自不同地区的晋谷21号小米进行溯源研究。方法使用近红外光谱仪获取产自洪洞、浮山、沁县3个不同地区的晋谷21号小米400~1004nm波段范围内的漫反射光谱;对光谱分别进行多元散射校正法(multiple scattering correction,MSC)、一阶导数法(first derivative,1St-D)预处理;对预处理光谱进行主成分分析,全交叉验证确定最佳主成分数量,获取主成分;同时选择预处理光谱特征波长。使用马氏距离法、线性判别法建立判别模型,最后用未知样品的验证准确率来表示模型的判别效果。结果原始光谱和MSC处理光谱提取特征波长分别建立的产地判别模型对3个不同产地的小米判别完全准确;1St-D处理光谱基于7个主成分结合马氏距离法和基于9个主成分结合线性判别法建立的2种判别模型对3个不同产地的小米亦实现完全准确判别。结论可见/近红外反射光谱技术用于小米产地的判别具有可行性,本研究可为小米产地的快速判别应用中提供技术基础。     

利用近红外光谱及电子鼻技术快速无损鉴别长期冻藏猪肉

利用可见-近红外光谱技术和电子鼻主成分分析技术对新鲜肉和长期冻藏猪肉分别进行快速无损鉴别.首先通过近红外分析技术采集新鲜肉样(-18℃分别贮藏0、3个月)和长期冻藏猪肉(-18℃分别贮藏6、9个月)的判别样本光谱图,对样品的光谱值进行直接判别分析,采用不同的预处理方式优化长期冻藏猪肉和新鲜肉的判别模型并分别进行主成分分...     

利用可见/近红外光谱判别干枣品种

目的利用可见/近红外反射光谱技术快速判别干枣的品种。方法使用光谱仪获取山西永和枣、山西板枣和新疆和田枣3种干枣在345~1100 nm波段范围内的漫反射光谱;分别使用多元散射校正(MSC)法和一阶导数法(1~(st)-D)和二阶导数法(2~(st)-D)对反射光谱进行预处理;对预处理光谱进行主成分分析,全交差验证法确定最佳主成分数量,提取主成分,结合马氏距离法和线性判别法建立品种判别模型,建立模型过程中使用全交叉验证法确定最佳主成分数,将模型应用于干枣的品种判别。结果可见/近红外反射光谱经过MSC处理后提取主成分建立品种预测模型对枣的品种判别结果最好,利用前4个主成分结合马氏距离法建立的判别模型和利用前5个主成分结合线性判别法建立判别模型,对于3个品种的枣的校正和验证判别准确率都达到了100%。结论可见/近红外反射光谱技术可以较好地判别干枣品种,本研究可为可见/近红外光谱技术在于枣品种和产地的快速鉴别和溯源中的应用提供一定的技术基础。     

基于近红外光谱的冷鲜肉--解冻肉的判别研究

目的 利用近红外光谱对冷鲜肉和解冻肉进行判别研究。方法 利用385～935nm的近红外光谱系统,采集冷鲜肉与解冻肉表面的反射光谱数据,采用Savitzky-Golay(S-G)平滑和标准变量正态变换(SNVT)方法对光谱数据进行预处理。然后利用主成分分析法(PCA)实现数据降维,提取主成分后结合两种状态肉的理化指标(L*值,a*值,pH,蒸煮损失和嫩度),分别利用Fisher判别法、贝叶斯判别法两种判别分析方法对冷鲜肉和解冻肉进行判别研究。结果 两种判别分析方法,均取得较好的分类效果,尤其是Fisher判别法,校正集的回判正确率为96.67%,验证集的正确率为100%。结论 近红外光谱技术应用于冷鲜肉和解冻肉的鉴别是可行的。     

基于电子鼻表征霉心病苹果特征气味及无损检测模型建立

为探究电子鼻检测技术对霉心病苹果的判别效果,以富士健康苹果和霉心病苹果为试材,基于SIMCA软件对采集的不同病变程度霉心病苹果的电子鼻信息进行表征,基于SPSS 23.0软件建立霉心病苹果Fisher函数、多层感知器神经网络(muhilayer perceptron neural network,MLPNN)和径向基函...     

基于近红外光谱的冷鲜肉-解冻肉的判别研究

目的利用近红外光谱对冷鲜肉和解冻肉进行判别研究。方法利用385~935 nm的近红外光谱系统,采集冷鲜肉与解冻肉表面的反射光谱数据,采用Savitzky-Golay(S-G)平滑和标准变量正态变换(SNVT)方法对光谱数据进行预处理。然后利用主成分分析法(PCA)实现数据降维,提取主成分后结合两种状态的肉的理化指标(L*值,a*值,pH,蒸煮损失和嫩度),分别利用Fisher判别法、贝叶斯判别法两种判别分析方法对冷鲜肉和解冻肉进行判别研究。结果两种判别分析方法,均取得较好的分类效果,尤其是Fisher判别法,校正集的回判正确率为96.67%,验证集的正确率为100%。结论近红外光谱技术应用于冷鲜肉和解冻肉的鉴别是可行的。     

Detection Potential of Multi-Features Representation of E-Nose Data in Classification of Moldy Maize Samples

In order to assess rapidly and timely the moldy degree of maize samples using electronic nose (E-nose) and improve the correct classification rate of E-nose, the different feature representation modes (DFRM) for E-nose data were explored in depth. A determining method for multi-features vector of E-nose based on Wilks Λ statistic was introduced so as to obtain the best multi-features vector for characterizing E-nose data. And then a selection method of representation features of each sensor signals based on elimination transform with pivoting of the Λ statistic was also introduced for the different excitation characteristic of each gas sensor. The research results show that the classification effect of multi-features representation mode (MFRM) is better than that of single feature representation mode (SFRM), and the MFRM is not a regular pattern, but the best multi-features vector of E-nose in MFRM can be obtained by the determining method. Moreover, it is necessary to select the representation features of each sensor signals in the MFRM using the selection method. The visual inspection results based on SFRM and MFRM were examined by Fisher discriminant analysis (FDA) and proved that the introduced methods were very effective, the highest correct discrimination rate based on SFRM is 80%, while the correct discrimination rate of the five features combination is 97%. As an outlook, we believe that the research findings may be universally applied for the classification of other food and agriculture products using E-nose.     

Authentication of pure camellia oil by using near infrared spectroscopy and pattern recognition techniques

Abstract: Total of 4 pattern recognition methods for the authentication of pure camellia oil applying near infrared (NIR) spectroscopy were evaluated in this study. Total of 115 samples were collected and their authenticities were confirmed by gas chromatography (GC) in according to China Natl. Standard (GB). A preliminary study of NIR spectral data was analyzed by unsupervised methods including principal component analysis (PCA) and hierarchical cluster analysis (HCA). Total of 2 supervised classification techniques based on discriminant analysis (DA) and radical basis function neural network (RBFNN) were utilized to build calibration model and predict unknown samples. In the wavenumber range of 6000 to 5750 cm^?1, correct classification rate of both supervised and unsupervised solutions all can reach 98.3% when smoothing, first derivative, and autoscaling were used. The good performance showed that NIR spectroscopy with multivariate calibration models could be successfully used as a rapid, simple, and nondestructive method to discriminate pure camellia oil.     

基于近红外光谱的豆酱鉴别技术研究

针对葱伴侣、凤彩桥、海天和金菜地四种品牌的豆酱,利用近红外光谱分析技术,对其进行预处理、主成分分析(PCA)、聚类分析(CA)和判别分析(DA),以建立识别不同豆酱品牌的近红外光谱定性判别模型.分析结果显示4种不同品牌的平均近红外光谱存在差异,其主成分空间分布也处于不同区域.对样品进行聚类分析,凤彩桥和海天存在小部分交...     

A selection method for feature vectors of electronic nose signal based on wilks Λ–statistic

There are mainly two selection methods for different features of electronic nose (E-nose) which is used to identify different samples, namely visual inspection and correct rate of discrimination result. The visual inspection is not a quantitative method. Besides, when the correct rates of discrimination result are identical for different features, the identification difference of different features is not evaluated accurately and quantitatively. To get a better feature vector for identifying different samples, a selection method was studied in-depth in which Wilks Λ–statistic was employed as a selection index for different features. At the same time, three different kinds of Chinese vinegar and three of Chinese milk were taken and tested by an E-nose. Five different features were extracted from the E-nose signals which are variance value (VARV), integral value (INV), mean value of relative steady-state responses (MVRSR), mean-differential coefficient value (MDCV) and energy value of wavelet packet decomposition (WE). The best feature vectors of these five features were obtained using the selection method and its effectiveness was respectively proved by the visual inspection and Fisher discriminant analysis (FDA) correct rate of vinegar and milk samples.     

Fast Discrimination of Apple Varieties Using Vis/NIR Spectroscopy

We evaluated the potential of visible/near-infrared (Vis/NIR) spectroscopy for its ability to nondestructively differentiate apple varieties. The apple varieties used in this research included, Fuji apples, Red Delicious apples, and Copefrut Royal Gala apples. The chemometrics procedures applied to the Vis/NIR data were principal component analysis (PCA), wavelet transform (WT), and artificial neural network (ANN). The apple varieties could be qualitatively discriminated in the PC1-PC2 space resulted from PCA. Wavelet transform was used as a tool for dimension reduction and noise removal, reducing spectral to wavelet components. Wavelet components were utilized as input for three-layer back propagation ANN model. WT-ANN model gave the highest level of correct classification (100%) of the apple varieties.     

Application of Gas Chromatography-Based Electronic Nose for Classification of Chinese Rice Wine by Wine Age

The potential of a gas chromatography (GC)-based electronic nose (E-nose) combined with chemometrics to classify Chinese rice wine by wine age was investigated. Olfactive fingerprints of 1-, 3-, and 5-year Chinese rice wine samples collected by the E-nose were analyzed by principal component analysis (PCA) and discriminant analysis (DA). Gas chromatography/mass spectrometry (GC/MS) was used for wine age validation purpose. The results indicated that the percentage of samples correctly classified by the E-nose was 96.88 %. It was concluded that the GC-based E-nose together with DA was a reliable method for wine age discrimination.     

A nondestructive approach for discrimination of the origin of sesame seeds using ED-XRF and NIR spectrometry with chemometrics

An energy dispersive X-ray fluorescence (ED-XRF) spectrometer and a near infrared (NIR) spectrometer combined with chemometrics were applied for origin discrimination of 48 Korean, 44 Chinese, and 21 Indian sesame seed samples used for development of a discriminant calibration model. Multi-elemental ED-XRF analysis based on Mg, Al, Si, P, S, Cl, K, Ca, Mn, Fe, and Cu was used for comparisons among origins. All elements, except for Fe, showed differences and 96.5% of seed samples were assigned to the correct origin using discriminant analysis based on chemical analytical results. NIR measurements were performed for spectral scanning. Classification of seeds using NIR discriminant analysis achieved 89.4% of seed samples assigned to the correct origin. Both ED-XRF and NIR are useful as nondestructive tools for discrimination of sesame seed origins.