基于NIR高光谱成像技术的滩羊肉内部品质无损检测

利用近红外高光谱成像技术对滩羊肉蛋白质和脂肪含量、pH值进行无损检测研究。通过高光谱系统(900~1700 nm)采集69个羊肉样本信息,先对全波段下的原始光谱和预处理后光谱建立偏最小二乘回归(PLSR)模型,对比优选出最佳预处理算法,后采用PLSR的加权β系数法提取特征波长,建立特征波长下各品质参数的PLSR模型,分析预测效果。结果表明:羊肉蛋白质、脂肪含量、pH值最佳预处理方法为基线校准(Baseline)、多元散射校正与S-G卷积平滑结合算法(MSC+SG)和原始光谱;利用特征波长建立预测模型,决定系数(RP2)分别为0.83、0.86和0.72,预测均方根误差(RMSEP)为0.57、0.09和0.12,可替代全波段建模。利用近红外高光谱成像技术对羊肉内部品质进行快速无损检测是可行的。     

冷鲜羊肉品质的高光谱成像无损检测

利用400¹000 nm可见近红外高光谱成像系统对冷鲜羊肉蛋白质含量、嫩度、p H进行无损检测研究。采集冷鲜羊肉表面的高光谱散射图像,提取样本感兴趣区域的反射光谱曲线获得原始数据。先对原始光谱预处理并建立偏最小二乘回归（PLSR）模型,优选最佳预处理方法,后采用正自适应加权算法（CARS）和连续投影算法（SPA）提取特征波长,建立不同特征波长下各品质参数的PLSR预测模型。结果表明:利用原始光谱建立的冷鲜羊肉蛋白质、嫩度和p H的PLSR模型均优于经过光谱预处理所建PLSR模型;在不同波长下建立预测模型,OS-PLSR光谱模型对冷鲜羊肉蛋白质含量预测效果最佳,Rp=0.869,RMSEP=0.097;建立的SPA-PLSR光谱预测模型对p H预测效果理想,Rp=0.958,RMSEP=0.067;CARS-PLSR光谱预测模型对嫩度的预测能力较高,Rp=0.862,RMSEP=0.706。研究表明:利用可见近红外高光谱技术对冷鲜羊肉品质进行快速无损检测是可行的。      

基于近红外光谱技术检测苹果气调贮藏期可溶性固形物含量

目的 基于近红外光谱技术结合偏最小二乘（Partial least square, PLS）法和最小二乘支持向量机回归（Least square-support vector regression, LS-SVR）法建立苹果气调贮藏期可溶性固形物（Soluble solids contents, SSC）含量预测模型。方法 在分析了气调贮藏期苹果细胞结构和SSC变化的基础上,采集了可见-近红外（Visible-near infrared, Vis-NIR）波段和长波近红外（Long wave near infrared, LWIR）波段下不同贮藏时间的苹果漫反射光谱信息,利用主成分分析方法（Principal component analysis,PCA）分析不同贮藏期苹果光谱信息分布特征,使用Kennard-Stone（K-S）算法以3:1比例对样本集进行划分,使用多元散射校正（Multiplicative scatter correction, MSC）和SG（Savitzky-Golay）平滑对光谱进行预处理,利用连续投影算法（Successive projections algorithm, SPA）和竞争自适应重加权采样（Competitive adaptive reweighted sampling, CARS）法对光谱进行特征波长提取,并建立SSC预测模型。结果 在LWIR波段下,经MSC预处理和CARS提取特征波长后建立的PLS模型取得了较好的预测精度,模型相关系数为0.900,均方根误差为0.478;经MSC、SG平滑预处理和CARS提取特征波长后建立的LS-SVR模型取得了更好的预测精度,模型相关系数为0.927,均方根误差为0.507。结论 构建的基于可见/近红外光谱无损预测模型可实现对气调贮藏期苹果SSC的准确预测,为高效贮藏技术提供了理论基础。     

生鲜猪肉水分含量的快速无损检测

目的 研究生鲜猪肉水分含量与1000~1680 nm范围内近红外吸收光谱之间的关系, 对生鲜肉的水分含量进行快速无损检测。方法 将原始光谱经中值平滑、多元散射校正和一阶导数复合预处理, 结合多元线性回归和偏最小二乘回归两种建模方法建立生鲜肉水分含量的预测模型。结果 应用所建立的模型对111个实际生鲜猪肉样品的水分含量进行预测, 得到较为满意的预测结果, 两种模型的预测相关系数分别为0.839和0.810。结论 所建模型适合于生鲜猪肉水分的无损快速检测。     

基于高光谱图像特征融合的榛子水分含量测定

采用高光谱图像技术对榛子水分含量进行快速无损检测。采集200个榛子在400~1 000 nm波段的高光谱图像,提取榛子图像区域的平均光谱信息。利用K-S算法划分样品验证集和预测集,使用四种预处理方法对光谱进行预处理。通过竞争自适应加权算法（Competitive Adaptive Reweighted Sampling,CARS）和逐次投影法（Successive Projection Algorithm,SPA）进行光谱特征的提取;灰度共生矩阵法（Gray Level Co-occurrence Matrix,GLCM）提取图像的纹理特征;分别建立基于光谱特征,图像纹理特征以及两者串联融合的偏最小二乘回归（Partial Least Squares Regression,PLSR）和支持向量回归（Support Vector Regression,SVR）模型对榛子水分进行预测。结果表明,CARS和SPA算法能够有效选择特征波长并提升预测性能;图像特征能够对榛子水分进行预测,基于主成分图像提取的图像特征信息建立的模型预测效果更好。光谱图像特征融合能明显提高对榛子水分含量预测的准确率,CARS提取的特征波段结合主成分图像的纹理特征显示出了更好的效果,SVR模型的RMSECV为0.03,RC 为0.97,RMSEP为0.04,RP为0.96。利用高光谱图像和纹理特征能够对榛子水分进行有效预测,为榛子水分含量检测提供了新的方法。     

盐池滩羊肉脂肪含量的高光谱预测模型构建

本文利用可见-近红外高光谱成像技术预测冷鲜滩羊肉脂肪含量,优选最佳预测模型。测定90个滩羊背最长肌的脂肪含量并采集其光谱图像,对原始光谱进行不同种预处理后,构建了全波段下的偏最小二乘回归(PLSR)和主成分回归(PCR)的光谱预测模型。为减少模型运算次数,在预处理效果最优的全波段模型上采用连续投影算法(SPA)、应用竞争性自适应重加权(CARS)、变量组合集群分析(VCPA)和波长空间迭代收缩(IVISSA)方法提取特征波长,构建脂肪含量的光谱预测模型。结果表明:采用归一化(Normlize)预处理后光谱构建的PLSR全波段模型效果最好,校正集模型相关系数(Rc)达到0.921;采用多元散射校正(MSC)预处理后光谱构建的PCR全波段模型效果最好,其校正集模型相关系数(Rc)达到0.850;在4种提取特征波长过程中,IVISSA算法所提取特征波长的交互验证均方根误差(RMSECV)最低,为0.0072;Normlize-IVISSA-PLSR模型较其他3种算法所构建的PLSR模型效果最优,其校正集相关系数(Rc)和预测集相关系数(Rp)值分别为0.931和0.754,表明利用高光谱技术对盐池滩羊肉脂肪含量进行预测是可行的。研究成果为冷鲜滩羊肉品质在线光谱快速无损检测系统开发提供理论依据。     

基于高光谱成像技术和连续投影算法检测葡萄果皮花色苷含量

应用高光谱成像技术结合连续投影算法（SPA）实现葡萄果皮中花色苷含量的快速无损检测。采集60 组样本高光谱图像,获取样本光谱曲线,并采用多元散射校正预处理方法提高信噪比。然后采用SPA选择光谱变量,将其作为多元线性回归（MLR）、偏最小二乘（PLS）模型和BP神经网络（BPNN）的输入变量,分别建立SPAMLR、SPA-PLS和SPA-BPNN模型并与全光谱变量PLS模型相比较。结果表明,SPA-MLR、SPA-BPNN和SPA-PLS模型的预测精度均优于全光谱变量PLS模型,其中SPA-PLS模型获得了最佳预测结果,其预测相关系数Rp和预测均方根误差（RMSEP）分别为0.900 0和0.550 6。结果表明,利用近红外高光谱成像技术能够有效检测酿酒葡萄果皮中花色苷含量。     

冷鲜羊肉冷藏时间和水分含量的高光谱无损检测

利用可见-近红外高光谱成像技术对冷鲜羊肉的冷藏时间和水分含量进行无损检测。通过波长400～1 000 nm可见-近红外高光谱系统采集160 个羊肉样本光谱信息,优选主成分-14-线性判别法对原始光谱建立羊肉冷藏时间的判别模型,校正集对羊肉冷藏时间的判别率为99.17%,预测集为100%,模型可较好地判别羊肉的冷藏时间。其次,针对羊肉冷藏过程中水分含量的变化,优选最佳预处理方法并运用偏最小二乘回归（partial leastsquares regression,PLSR）法建立水分含量预测模型;结果表明,经过Savitzky-Golay卷积平滑预处理的PLSR模型对水分含量的建模效果最优,校正集和预测集相关系数分别为0.888和0.784,交互验证均方根误差为0.696。研究表明,采用可见-近红外高光谱成像技术对冷鲜羊肉冷藏时间的判别和冷藏过程中羊肉水分含量的快速预测是可行的。     

基于可见/近红外光谱预测枇杷糖度及模型优化

为实现枇杷糖度的快速无损检测,并探究开阳枇杷糖度最优预测模型。首先利用光纤光谱仪获取开阳枇杷的反射光谱,分析比较标准正态变换和多元散射校正方法对原始光谱数据的预处理效果;然后基于原始全光谱和预处理后的全光谱数据分别构建预测开阳枇杷糖度的偏最小二乘回归和主成分回归模型;最后,采用连续投影算法和竞争性自适应重加权算法（competitive adaptive reweighted sampling, CARS）筛选特征光谱,并基于特征光谱构建预测开阳枇杷糖度的多元线性回归（multi linear regression, MLR）模型。结果表明,采用CARS算法从785个全光谱中筛选了23个特征波长,不仅提升了预测模型的运算效率,而且建立的CARS-MLR模型具有最佳的校正性能（R_C=0.89,RMSEC=0.62）和预测性能（R_P=0.89,RMSEP=0.65,RPD=2.29）。这表明利用可见/近红外光谱技术结合化学计量学预测开阳枇杷糖度是可行的,且CARS-MLR模型相对最优,为枇杷品质的无损快检和分选提供了理论依据与技术基础。     

NIR高光谱成像技术检测冷鲜羊肉嫩度

利用900¹700 nm近红外高光谱成像系统对冷鲜羊肉嫩度进行快速无损检测研究。采集冷鲜羊肉（1⁸ d）表面的高光谱散射图像,提取样本感兴趣区域反射光谱曲线并用剪切力值表征冷鲜羊肉的标准嫩度。以原始光谱、特征区域光谱和Savitzky-Golay卷积平滑预处理光谱建立冷鲜羊肉嫩度的偏最小二乘回归（PLSR）模型,预处理的特征区域光谱建立的模型效果更优。结果表明:特征区域光谱可有效替代全波段光谱,经过S-G卷积平滑预处理后,模型预测效果最佳,预测相关系数（Rp）和均方根误差（RMSEP）分别为0.773和1.060。研究表明:利用近红外高光谱成像技术结合偏最小二乘回归法对冷鲜羊肉嫩度的快速无损检测是可行的。      

宁夏赤霞珠葡萄水分含量的高光谱无损检测研究

利用可见近红外高光谱成像技术对宁夏赤霞珠葡萄含水量的无损检测进行了初步探讨。通过高光谱成像系统(400~1000 nm)采集了136幅赤霞珠葡萄图像,对原始光谱、平均平滑、高斯滤波、中值滤波、卷积平滑、归一化、多元散射校正、标准正态化、基线校准、去趋势化等预处理的偏最小二乘回归(PLSR)模型进行对比分析;采用主成分分析(PCA)、偏最小二乘回归(PLSR)、连续投影算法(SPA)、竞争性自适应重加权(CARS)方法选择特征波长,建立4种特征波长下的PLSR的葡萄含水量预测模型,优选CARS提取特征波长的方法。在此基础上,对比分析了全波段与特征波长下的MLR、PCR、PLSR的葡萄含水量预测模型。结果表明:采用多元散射校正(MSC)光谱建立的PLSR模型优于原始光谱和其他预处理光谱的PLSR模型;CARS提取特征波长建立的PLSR模型优于多元线性回归(MLR)、主成分回归(PCR)模型,预测集的相关系数(R)和预测均方根误差(RMSEP)分别为0.806、0.144。因此,利用可见近红外高光谱成像技术提取特征波长进行宁夏赤霞珠葡萄含水量的检测是可行的。      

基于高光谱成像技术的金银花与山银花快速鉴别

利用高光谱成像技术,研究一种快速、准确、无损的鉴别金银花与山银花的方法。通过对比3种预处理方法对偏最小二乘算法(Partial Least Squares,PLS)建模效果的影响,得到SNV为建模最优预处理方法。使用回归系数法(Regression Coefficient,RC)和连续投影算法(Successive Projection Algorithm,SPA)选择经预处理后光谱的特征波长,并分别建立极限学习机(Extreme learning machine,ELM)和最小二乘支持向量机(Last Squares Support Vector Machine,LSSVM)的判别分析模型。结果表明,光谱经SNV预处理后,应用SPA提取特征波长并建立LS-SVM判别分析模型为金银花和山银花最优判别模型,其建模集与预测集识别率均达到了100.00%。因此,利用高光谱成像技术能够无损、有效地鉴别金银花与山银花,并且在全光谱和特征波长下均能实现金银花与山银花的快速判别分析。     

Analysis of protein denaturation,and chemical visualisation,of frozen grass carp surimi containing soluble soybean polysaccharides

The mechanism of protein denaturation of frozen surimi enriched with soluble soybean polysaccharides (SSPS) was investigated. Near-infrared (NIR) and hyperspectral imaging (HSI) technology were used to predict protein denaturation. The fresh grass carp surimi was divided into four groups with SSPS additions of 0%, 1%, 3% and 5%, respectively, which were frozen and stored at −18 °C. Samples were examined after 0, 1, 2, 4 and 8 weeks, for salt-soluble protein content, total SH, NIR hyperspectral image and Raman spectrum features. The results showed that addition of SSPS decelerated protein degradation. After 8 weeks of storage, the 5% SSPS addition maintained the highest salt-soluble protein content, while for total SH, 3% and 5% SSPS addition had similar effect. Raman spectra illustrated that SSPS had the ability of maintaining an α-helix content, as well as decreasing the exposure of polar groups, and reducing the oxidation of SH group into disulphide bonds. NIR spectra showed that the overall reflectance of frozen surimi increased with the increase in SSPS. Two partial least squares regression (PLSR) models were established after pretreatment of HSI spectral statistics and selection of characteristic wavelengths. The distribution maps were finally generated based on the simplified PLSR models.     

白酒糟醅中酸度值的高光谱检测方法

酸度值是评价白酒糟醅质量的重要指标之一,为进一步提高糟醅酸度值的检测精度,提出了一种应用高光谱成像技术检测糟醅酸度值的方法.采用高光谱成像系统,在900～1700 nm内采集糟醅样本的光谱信息,并提取全部样本的平均光谱数据.采用3种预处理方法对原始光谱进行预处理,得到多元散射校正(multiplica-tive sca...     

Data Mining-Based Wavelength Selection for Monitoring Quality of Tomato Fruit by Backscattering and Multispectral Imaging

The aim of this research was to predict quality factors of tomato fruit during storage using backscattering and multispectral imaging techniques. To gather the required information for developing prediction models, batches of 200 tomatoes (cv. Pannovy) harvested at two maturity stages, were stored at standard condition up to four weeks. During storage, the modulus of elasticity, moisture content, soluble solid content, titratable acidity, hyperspectral data, and backscattering images were acquired on 40 tomatoes at regular intervals of one week. After extracting the spectral data from 40 points on each sample, they were subjected to preprocessing operations. Several feature selection techniques, including filter (Relief F, Fisher-Score, and t-Score) and wrapper (genetic algorithm) methods were used to find the sensitive wavelengths for each fruit quality parameter. With the novel strategy used, the wavelengths found by the fusion of genetic algorithm and t-Score techniques showed good prediction performance for all considered qualitative parameters. In order to verify the usefulness of selected wavelengths, backscattering and multispectral imaging techniques were applied. The artificial neural network produced the calibration models which gave a reasonably good correlation for estimating the modulus of elasticity, soluble solid content, and titratable acidity at 660 nm and moisture content at 830 nm of tomato from backscattering images. The correlation coefficient between the multispectral and backscattering imaging prediction results and reference measurement results were 0.952 and 0.891 for modulus of elasticity, 0.727 and 0.539 for moisture content, 0.736 and 0.561 for soluble solid content, and 0.811 and 0.706 for titratable acidity, respectively.     

Shortwave infrared hyperspectral imaging for detecting sour skin (Burkholderiacepacia)-infected onions

Sour skin (Burkholderiacepacia) is a major postharvest disease for onions and causes substantial production and economic losses in onion postharvest. In this study, a shortwave infrared hyperspectral imaging system was explored to detect sour skin. The hyperspectral reflectance images (950-1650 nm) of onions were obtained for the healthy and sour skin-infected onions. Principal component analysis conducted on the spectra of the healthy and sour skin-infected onions suggested that the neck area of the onion at two wavelengths (1070 and 1400 nm) was most indicative of the sour skin. Log-ratio images utilizing the two optimal wavelengths were used for two different image analysis approaches. The first method applied a global threshold (0.45) to segregate the sour skin-infected areas from log-ratio images. Using the pixel number of the segregated areas, Fisher’s discriminant analysis recognized 80% healthy and sour skin-infected onions. The second classification approach used three parameters (max, contrast, and homogeneity) of the log-ratio images as the input features of support vector machine (Gaussian kernel, γ = 1.5), which discriminated 87.14% healthy and sour skin-infected onions. The result of this study can be used to further develop a multispectral imaging system to detect sour skin-infected onions on packing lines.     

Moisture sorption studies on onion powder

Moisture sorption studies were conducted on three samples of onion powders, i.e. freeze-dried, vacuum shelf-dried and through flow-dried onions. The data obtained for texture (flowability) and per cent moisture were used in determining critical moisture levels. The monolayer moisture contents, corresponding to Brunauer Emmett Teller (BET) theory were taken as the safe minimum moisture levels in onion powder. The moisture content values (moisture-free basis), which correspond to a monomolecular layer of adsorbed water computed by BET equation, were 2.09, 1.96 and 1.94%, for freeze-dried, vacuum shelf-dried and through flow-dried onion powders, respectively. This study on initial and critical moisture levels of dehydrated onion powders helps in the selection of the packaging materials for storage.     

基于高光谱成像的干燥胡萝卜片水分及类胡萝卜素含量无损检测和可视化分析

利用不同波长范围的高光谱成像系统,以热风干燥过程中胡萝卜片水分和类胡萝卜素含量为研究对象,结合多元数据统计分析和化学计量学,分别构建基于偏最小二乘和支持向量机(support?vector?machine,SVM)算法的无损预测模型,并进行可视化分析。结果表明,水分和类胡萝卜素含量预测模型中,基于400～1?000?nm波长范围下多元散射校正的高光谱信息构建的SVM预测模型效果相对最优,对应的预测集决定系数R2P分别为0.984和0.911,预测集均方根误差(root?mean?square?error?of?prediction,RMSEP)分别为0.380?g/g和34.836?mg/100?g。经连续投影算法提取特征波长后,最优模型R2P分别为0.962和0.898,RMSEP分别为0.612?g/g和37.544?mg/100?g,模型剩余预测残差均大于3,精确度和稳定性良好。在最优预测模型的基础上,通过伪彩色成像重现了干燥过程中样品水分及类胡萝卜素的空间分布。实验结果证实高光谱成像技术可以用于胡萝卜片干燥过程水分和类胡萝卜素含量的无损检测,为后续在线检测和胡萝卜片干燥加工提供...     

基于高光谱成像技术无损检测芒果轻微损伤

目的 使用高光谱成像技术实现对芒果轻微损伤的无损识别。方法 在可见光-近红外波长范围内采集完好芒果和损伤芒果的高光谱图像,并提取相应的感兴趣区域(regions of interest, ROI)获得样本高光谱数据。经过多种预处理方法比较,选择光谱预处理方法。使用竞争性自适应重加权算法(competitiveadaptivereweighted sampling, CARS)和连续投影算法(successive projections algorithm, SPA)分别对预处理后的光谱提取特征波长,并分别建立了多元线型回归(multiplelinearregression,MLR)模型和偏最小二乘回归(partialleastsquaresregression,PLSR)模型。结果 选择多元散射校正(multiplicative scatter correction, MSC)作为光谱预处理方法。针对芒果轻微损伤识别,CARS-MLR模型识别效果最好,其校正集相关系数为0.881,预测集相关系数为0.821,校正集均方根误差(calibration set root mean squa...