小麦子粒粗蛋白FT-NIRS分析模型建立的初步研究

采用凯氏定氮法测试了216份小麦子粒的粗蛋白含量,用近红外仪采集数据,选择113份建立了数学模型。结果:最佳主成分数(Rank)=6,内部交叉验证均方差(RMSECV)=0.377,决定系数(R2)=96.89。为了验证模型的可靠性,对预测集样品进行预测,结果粗蛋白的预测均方差(RMSEP)=0.950,相对偏差(RSEP,%)=8.40。     

应用近红外反射光谱技术测定油菜芽期耐旱性的研究

针对油菜芽期耐旱鉴定，提出了用近红外反射光谱技术（NIR法）预测油菜吸胀24 h电导率、PEG模拟干旱条件下的相对发芽率、相对鲜重和鲜重耐旱指数等4个芽期耐旱相关性状的方法。以采集的49份不同耐旱水平甘蓝型油菜近红外光谱数据为基础，采用偏最小二乘法和多元回归算法建立了最优定标模型，并获得较高的决定系数(0.71～0.86)和较低的标准误差(1～15.65)。验证集评估结果表明，NIR法与室内鉴定法测定油菜4个芽期耐旱相关性状无显著差异，且具有极显著的相关关系(决定系数0.72～0.89)。研究表明,近红外光谱技术用于油菜芽期耐旱性鉴定是可行的，可用于耐旱育种早代选择。     

基于田间原位土壤含水量估测的可见/近红外光谱建模方法

为实时、准确地获取原位土壤含水量信息,利用可见/近红外光谱技术,分别使用全局偏最小二乘(PLS)建模、局部PLS建模方法,对田间原位土壤含水量进行快速估测。结果表明:全局PLS模型中,其建模集的决定系数(R~2)、交叉验证均方根误差(RMSECV)分别为0.943和1.750%,检验集的决定系数(R~2)、预测均方根误差(RMSEP)分别为0.956和1.260%。局部PLS模型中,分别比较了选取定标子集的2种方法(欧氏距离法和马氏距离法),采用欧氏距离法和马氏距离法选取定标子集进行建模的R~2值分别为0.974和0.979,RMSEP值分别为0.976%和0.943%。因此,将可见/近红外光谱技术应用到田间原位含水量测量是可行的,其中,使用局部建模方法的效果优于全局建模。     

巴山木竹蛋白质可见/近红外光谱定量分析模型研究

探讨了运用可见/近红外光谱分析技术建立巴山木竹蛋白质定量分析模型的可行性。运用传统方法实测了样品蛋白质含量,并运用光谱分析软件建立了样品蛋白质含量与光谱数据的PLS与PCR校正模型。基于主要性能指标对不同光谱预处理与建模方法进行评价,筛选出最优校正模型并使用验证集样品对校正模型的预测能力进行了验证。巴山木竹竹叶与竹秆蛋白质最优校正模型的决定系数(R_c~2)分别为0.935和0.862,交叉验证均方差(RMSEC)分别为0.351和0.172;经外部验证,预测模型决定系数(R_p~2)分别为0.916和0.874,验证集样品的相对分析误差(RPD)分别为3.562和2.840。表明应用可见/近红外光谱分析技术可以实现巴山木竹蛋白质含量快速检测。     

基于近红外光谱法的绿豆淀粉和蛋白质无损快速检测

为实现绿豆籽粒淀粉和蛋白质含量的无损快速检测,满足高淀粉、高蛋白绿豆的筛选、选育等需要,选用我国不同地区100份绿豆种质资源为材料,采用酸水解法和凯氏定氮法分别测定淀粉和蛋白含量,样品中淀粉含量范围为40.88%～53.62%,蛋白含量范围为20.17%～27.38%,化学测量值数据分布广。应用近红外光谱技术建立绿豆中主要营养物质的快速检测方法。通过采集样品的近红外光谱,结合化学方法所得数据建立近红外定量模型。结果表明,采用无光谱预处理可以得到最优的淀粉数学模型,内部交叉验证决定系数(r²)和交叉验证均方根误差(RMSECV)分别为0.926 9和0.658。采用一阶导数+MSC处理得到最优的蛋白模型,内部交叉验证决定系数r²和RMSECV分别为0.9 341和0.384。淀粉和蛋白外部验证决定系数r²分别为0.935 1和0.921 2,模型预测效果较好。建立的近红外模型可用于绿豆籽粒淀粉和蛋白质含量的无损快速检测,模型涵盖了绿豆种质淀粉和蛋白质含量的广泛检测分布范围,可实现绿豆资源的快速评价,提高育种效率。     

近红外光谱技术预测胡麻饼营养成分的定标模型建立及验证

【目的】探讨近红外光谱技术用于胡麻饼营养价值评定的可行性.【方法】以不同来源的胡麻饼为试验材料,在波长400~2 498nm,以改进最小二乘法搭配不同光谱预处理建立胡麻饼水分、粗灰分、粗蛋白、粗脂肪和总能的近红外定标模型.【结果】各指标模型交叉验证相关系数(1-VR)均在0.88以上,内部交叉验证集验标决定系数(RSQ)值均在0.92以上,外部验证集验标决定系数(RSQ)值均在0.85以上.【结论】近红外光谱技术能够快速无损测定胡麻饼营养成分,可以使胡麻饼更好地应用于动物饲料中.     

小批量稻谷种子蛋白质含量的近红外透射光谱分析

以完整水稻种子为样品,利用近红外透射谷物分析仪对186份批量稻谷进行扫描并测定了蛋白质含量的参比数据。采用多种数学计量学处理方法和不同的回归统计方法进行定标曲线的开发和比较,优化得到了小批量水稻种子蛋白质含量测定的近红外定标方程。其定标标准偏差(SEC)、交叉检验标准误差(SECV)、定标相关系数(RSQ)和交叉验证相关系数(I-VR)分别为0.255 8、0.279 5、0.972 8、0.967 5。研究采用整粒小量样品(5 g)来分析,效果较好,可直接用于育种早世代选择。     

基于近红外光谱的小麦种子发芽率测试

以534份发芽率水平不同的小麦品种种子为样品,采用傅里叶变换近红外光谱仪采集光谱数据,利用偏最小二乘法(PLS)建立其发芽率的无损测定校正模型,并对模型进行留一法交叉验证、外部验证。结果表明,经一阶导数和多元散射校正(MSC)预处理后,对7 502.3~4 246.8 cm-1波段范围所建模型的预测性能最佳,校正集决定系数R2为0.914 4,校正均方根误差(RMSEE)为7.38,平均绝对误差为5.925%;验证集决定系数R2为0.904 4,验证均方根误差(RMSEP)为7.91,平均绝对误差为6.467%。近红外光谱与种子发芽率具有较高相关性,利用近红外光谱技术快速测定小麦种子发芽率具有可行性。     

小麦子粒粗蛋白FT-NIRS分析模型建立的初步研究

采用凯氏定氮法测试了216份小麦子粒的粗蛋白含量,用近红外仪采集数据,选择113份建立了数学模型。结果：最佳主成分数（Rank）=6,内部交叉验证均方差（RMSECV）=0.377,决定系数（R2）=96.89。为了验证模型的可靠性,对预测集样品进行预测,结果粗蛋白的预测均方差（RMSEP）=0.950,相对偏差（RSEP,%）=8.40。     

水稻蛋白质含量NIR模型适配范围的研究

【目的】比较不同类型样品建立水稻蛋白质近红外模型的效果和适配范围。【方法】通过对178份来自“II-32B/岳早籼6号”的重组自交系和496份水稻品种的近红外反射光谱的比较分析，选择其中59个株系和76份品种作为建模样品，采用偏最小二乘法建立基于品种、重组自交系和混合样品的3个蛋白质含量回归模型。【结果】经模型内部交叉验证和对模型外部重组自交系和品种样品的验证结果的比较分析，发现基于分离群体的模型因蛋白质含量范围较窄，样品来源较单一，适应范围仅局限于本群体内样品蛋白质含量预测，而品种和混合模型对群体和品种样品都表现出良好的适应能力，交叉验证决定系数大于0.90，外部验证决定系数大于0.89, 本试验可为近红外建模的样本集选择提供良好的指导意义。【结论】不同类型样品对建模效果有显著影响，品种模型和混合模型的适配范围显著大于群体模型，研究结果不能支持用背景变异较小的样品建立较高精度回归模型的设想。     

利用近红外光谱技术预测杉木力学性质

利用三点弯曲实验方法测定了155个杉木样品的抗弯弹性模量和抗弯强度,并用近红外光谱仪采集其径切面和横切面的近红外漫反射光谱,以2/3的试样(103个)作为校正集建立抗弯弹性模量和抗弯强度的偏最小二乘法校正模型,以1/3的试样(52个)作为预测集对模型进行验证.结果表明,切面对模型预测效果的影响比较小,且与光谱区域的选择有关.对于可见近红外全波段光谱(350～2 500 nm)利用径切面比利用横切面光谱建立的力学性质模型的预测效果略好,对于短波光谱(780～1 050 nm)利用横切面比利用径切面光谱建立的模型的预测效果略好;降低波谱范围后,利用横切面短波近红外光谱建立的力学性质校正模型的效果与全波谱模型相比差异较小;利用径切面和横切面2个切面可见近红外全波段光谱、利用横切面短波光谱分别建立的杉木力学性质的校正模型,其预测相对分析误差在1.51～1.90之间,表明利用近红外光谱技术预测杉木木材的力学性质的能力属普通,可用之作为初步的检测工具.     

基于CARS-BPNN的江西省土壤有机碳含量高光谱预测

【目的】探讨光谱变量选择及依据土壤类型进行分层校准两种方法对高光谱预测土壤有机碳（SOC）精度的影响。【方法】以江西省为研究区,490个土壤样本为研究对象,对研究区内的所有样本以及不同土壤类型样本分别通过竞争性自适应重加权采样（CARS）算法筛选特征波段,并采用偏最小二乘回归（PLSR）、支持向量机（SVM）、随机森林（RF）、反向传播神经网络（BPNN）4种模型,对比不同土壤类型下SOC在全波段以及CARS算法筛选后特征波段的预测精度。进而,还对比了全局校准和分层校准下SOC在全波段以及CARS算法筛选后特征波段的预测精度。【结果】（1）红壤筛选的特征波段为484、683—714和2 219—2 227 nm,水稻土筛选的特征波段为484、689—702和2 146—2 156 nm。红壤采用CARS-BPNN模型预测效果最佳（R ²=0.82）,较全波段建模验证集R ²提升0.07。水稻土采用CARS-RF模型预测效果最佳（R ²=0.83）,较全波段建模验证集R ²提升0.13。（2）在总体样本上,分层校准相比全局校准精度有所提升。采用CARS-BPNN进行分层校准预测效果最佳（R ²=0.82）,较全局校准验证集R ²提升0.06。【结论】采用CARS-BPNN进行分层校准能够较好地预测江西省土壤有机碳含量,本研究可为其他类似地区预测土壤属性提供科学依据。     

Rapid Non-destructive Detection for Molds Colony of Paddy Rice Based on Near Infrared Spectroscopy

Near infrared spectrometer technology under a wavelength range of 918-1045 nm was used to rapidly detect paddy rice that was stored at 5℃, 15℃ and 25℃. A total of 121 paddy rice samples were collected from artificial infection with moulds to build the calibration models to calculate the total number colony of moulds based on the principal component regression method and multiple linear regression method. The results of statistical analysis indicated that multiple linear regression method was applicable to the detection of the total number colony of moulds. The correlation of calibration data set was 0.943. The correlation of prediction data set was 0.897. Therefore, the result showed that near infrared spectroscopy could be a useful instrumental method for determining the total number colony of moulds in paddy rice. The near infrared spectroscopy methodology could be applied for monitoring mould contamination in postharvest paddy rice during storage and might become a powerful tool for monitoring the safety of the grain.     

The potential of visible-near infrared hyperspectral imaging to discriminate between casing soil, enzymatic browning and undamaged tissue on mushroom (Agaricus bisporus) surfaces

The potential of hyperspectral imaging (HSI) in the visible-near infrared (445-945 nm) wavelength range to discriminate between casing soil, enzymatic browning and undamaged tissue on mushroom (Agaricus bisporus) surfaces was investigated. A calibration set of 108 damage free mushrooms, grown under controlled conditions in a research station, were first tested as undamaged class (U) and then were divided into 2 groups of 54 samples. The first group was smeared with casing soil and designated as casing soil class (C) and the second group was subjected to vibrational damage resulting in enzymatic browning and designated as damaged class (D). Partial least squares discriminant analysis (PLS-DA) models were developed to classify mushroom tissue as one of the three classes investigated (U, C and D) using pixel spectra from each class. Prediction maps were obtained by applying the developed models to the hyperspectral images of candidate mushrooms. Percentages of pixels classified into each class were also calculated for the mushrooms studied in the calibration set. Results obtained showed that the developed models performed satisfactorily to discriminate between the 3 classes studied. Comparison of red-green-blue (RGB) and hyperspectral image analysis showed that HSI was better able to identify the regions containing casing soil. Model validation was performed using 3 different test sets of mushrooms obtained from a commercial producer. It was found that the developed PLS-DA models were satisfactorily capable of identifying undamaged regions, casing soil and enzymatic damaged areas on mushrooms from the validation sets.     

长波近红外光谱在土壤总氮分析与稳定性中的应用

[目的]通过定标集、预测集、检验集的建模过程,采用偏最小二乘(PLS)方法结合波段选择建立土壤总氮快速分析的近红外(NIR)光谱模型。[方法]为了避免模型评价失真,基于随机性、相似性和稳定性,提出一种严谨的建模体系。将全谱扫描区(400～2 498nm)分成可见区(400～780 nm)、短波近红外区(780～1 100 nm)和长波近红外区(1 100～2 498 nm)。[结果]经过比较、检验,结果表明长波近红外达到了最好的模型效果和稳定性,最优PLS因子数为8,检验的预测均方根误差(V-SEP)和预测相关系数(V-RP)分别为0.118 g/kg和0.857,得到客观、稳定的预测模型。     

基于样本挑选和不同偏最小二乘方法的近红外光谱玉米淀粉组分校正模型的研究

[目的]获得精度高、鲁棒性强的玉米近红外光谱淀粉组分检测模型。[方法]用一阶导数和Savitzky．Golay平滑对玉米1300～2298nlTl近红外光谱进行预处理,而后分别以RS（random sampling）、KS（Kennard Stone）、Duplex、SPXY（sample set partitioning based on joint x-y distance）方法选取最佳校正集样本集合,最后分别用PLS（Partial Least Squares）、iPLS（intervalPLS）和siPLS（synergy interval PLS）方法建立校正模型。[结果]采用sPXY方法选取有代表性的校正集合样本,以siPLS方法所建立的近红外光谱玉米淀粉组分校正模型最优,校正样本集合中r为0．9917,RMSECV为n1073,预测样本集合中r达到了0．9944,RMSEP为0．0814。[结论]SPXY-siPLS方法建立的近红外光谱玉米淀粉组分校正模型,不但可以减小参与建模的数据规模．而且缩短了运算时间．预测能力和精度也均得到提高。     

基于可见-近红外光谱技术的油菜叶片叶绿素含量无损检测研究

    为了快速无损获取油菜叶片叶绿素含量信息,试验研究了油菜叶片的可见-近红外反射光谱特性与叶绿素含量之间的定量关系.试验采集140个油菜叶片样本,其中70个样本用于建模,另外70个样本用于模型预测.光谱曲线扫描采用美国USB4000光纤光谱仪,叶绿素含量值采用日本Minolta 公司生产的SPAD-502仪测定.实验发现,波段范围680～730 nm处的光谱吸光度与油菜叶片叶绿素含量之间具有显著相关性.同时发现油菜叶片厚度对建模预测精度有较大影响.试验首先用待定系数法构造叶绿素含量预测方程;然后用标准遗传算法对其进行参数优化.试验确定最优光谱范围是696.82～716.53 nm.不考虑叶片厚度时,建模和预测关联度r分别是0.4823 和0.5649.考虑叶片厚度校正后,建模和预测关联度r分别提高到0.8936 和0.9178.说明基于可见-近红外反射光谱技术实现油菜叶片叶绿素含量快速无损检测是可行的.     

The use of near infrared (NIR) spectroscopy to improve soil mapping at the farm scale

The creation of fine resolution soil maps is hampered by the increasing costs associated with conventional laboratory analyses of soil. In this study, near infrared (NIR) reflectance spectroscopy was used to reduce the number of conventional soil analyses required by the use of calibration models at the farm scale. Soil electrical conductivity and mid infrared reflection (MIR) from a satellite image were used and compared as ancillary data to guide the targeting of soil sampling. About 150 targeted samples were taken over a 97 hectare farm (approximately 1.5 samples per hectare) for each type of ancillary data. A sub-set of 25 samples was selected from each of the targeted data sets (150 points) to measure clay and soil organic matter (SOM) contents for calibration with NIR. For the remaining 125 samples only their NIR-spectra needed to be determined. The NIR calibration models for both SOM and clay contents resulted in predictions with small errors. Maps derived from the calibrated data were compared with a map based on 0.5 samples per hectare representing a conventional farm-scale soil map. The maps derived from the NIR-calibrated data are promising, and the potential for developing a cost-effective strategy to map soil from NIR-calibrated data at the farm-scale is considerable.     

Multispectral imaging system with interchangeable filter design

The design and calibration of a three-band image acquisition system was reported in this paper. The prototype system developed in this research was a three-band spectral imaging system that acquired two visible-band images and one NIR image simultaneously. This was accomplished by using a three-port imaging system that consisted of three identical monochrome cameras, an optical system, and three interchangeable optical filters. Spectral reflectance from an object was collimated by a front lens, and split in three ways by a cold mirror and beamsplitter: a cold mirror reflects 90% visible light and transmits 80% NIR light. The visible light was again split identically into two directions by an additional beamsplitter. Focusing lenses then projected each image onto its corresponding sensor. By incorporating an interchangeable filter design, the imaging system can measure any two visible spectral bands that range between 400 nm and 700 nm, and one NIR band that ranges between 700 nm and 1000 nm without any complicated manufacturing process. In order to co-register the three images, a system-specific calibration algorithm was developed that compensates for lens-sensor geometric misalignments.The prototype imaging system and the system calibration algorithm were tested and evaluated for image alignment accuracy. The imaging system acquired three-band images of 3D objects with 0.39 pixel misalignment error on average.     

近红外光谱法预测粗皮桉木材的化学成分质量分数

采用近红外光谱法对粗皮桉木材中化学成分质量分数进行快速预测。用常规湿化学方法测定了粗皮桉木材样品的化学成分质量分数,结合近红外光谱仪采集相应的光谱,对原始光谱进行二阶导数预处理后,用偏最小二乘法建立相应的模型并对其进行外部验证。结果表明:粗皮桉木材综纤维素校正模型的相关系数为0.96,预测模型的相关系数为0.92,RPD为2.30。木质素校正模型的相关系数为0.91,预测模型相关系数为0.88,RPD为2.11。利用近红外光谱分析方法可以快速预测粗皮桉木材中综纤维素和木质素质量分数。