基于光谱特征参数的温室番茄叶片叶绿素含量预测

为了快速、准确估测温室番茄叶片叶绿素含量,提升作物精细管理水平,利用光谱分析技术研究了温室番茄不同生长阶段叶绿素含量和响应光谱的相关性,在幼苗营养生长阶段叶片叶绿素含量呈增长趋势,到移植50天前后达到最大值,在此期间反射光谱的红边会向红外方向（长波）偏移,同时绿峰向蓝光（短波）方向偏移,绿峰幅值减小。从结果期开始叶绿素含量呈下降趋势,而红边、绿峰及绿峰幅值向相反方向变化。为了定量分析叶绿素含量和叶片反射光谱间的关系,从自定义的68个光谱特征参数中提取了7个能反映叶绿素含量变化的最优参量,并使用逐步回归、岭回归、主成分分量回归和偏最小二乘回归消除了最优参量的多重共线性,建立了叶绿素含量预测模型,其中岭回归模型精度最佳,均方根误差（RMSE）为0.406,决定系数(R2)为0.839。     

基于光谱技术的褐壳血斑蛋鉴别方法研究

血斑蛋是一种带有血丝的异常蛋。通过自制的鸡蛋内部品质光谱检测系统,采集40个正常蛋和60个人工注射血样的血斑蛋的可见/近红外光谱,研究比较了3种不同的血斑蛋判别方法:传统的血值判别法、偏最小二乘判别法(DPLS)以及融合光谱信息与蛋壳颜色信息的最小二乘支持向量机(LS-SVM)判别法,结果表明基于颜色信息融合的最小二乘支持向量机的判别结果明显优于传统的血值判别法,正常蛋的判别正确率为90%,血斑蛋的判别正确率为91.7%,证明了此方法可用于褐色蛋的血斑检测。     

基于组合核函数的籼稻重度不宜存检测模型

为给籼稻储存品质的判定提供一种快速无损检测手段,对80份重度不宜存籼稻和80份非重度不宜存籼稻的近红外反射光谱进行了实验研究。根据训练样本非线性可分的特点,选择支持向量机方法建立定性模型。在对不同核函数的特性进行分析和研究的基础上,定义了一种新的核函数——组合核函数。该组合核函数是多项式核函数与径向基核函数的线性组合,将两者各自的特点融合在一起兼具内推和外推性能。实验结果表明,以这两种函数的线性组合作为核函数且调节因子为0.7时,所建立的模型综合性能最好。所建模型的训练集正确识别率为97.21%,测试集正确识别率为93.25%。     

近红外光谱分析技术在木材机械性能检测中的研究进展

近红外光谱技术具有快速、无污染、成本低廉、准确性高等优点,相比其他无损检测方法,被广泛应用于农业、医学、化工、造纸等各个领域。国内外许多科研工作者在木材材性分析和检测方面也作了大量的探索。本文主要介绍木材材性分析的重要性、近红外光谱技术的基本原理和特点,国内外林业科技工作者在辐射松、蓝桉、火炬松和粗皮桉等不同树种木材的抗弯强度、抗弯弹性模量、密度和压缩强度等物理力学性质检测方面所做出的贡献和取得的进展,通过介绍可以看出NIR技术具有很大的潜能,它可以快速、准确的获悉木材的性质,从而对木材进行科学合理的利用。     

Modelling of pH effects and CIE L*a*b*colour spaces of beech wood-inhabiting fungi by NIRS

A combination of statistical techniques of analyses were used to evaluate the potential of International Commission on Illumination (CIE) lightness (L*), redness (a*) and yellowness (b*) colour space system and near-infrared spectroscopy (NIRS) to assess surface changes in relation with progressive decay of beech (Fagus grandifolia Ehrh.) by wood-inhabiting lignicolous fungi Inonotus hispidus, Trametes versicolor and Xylaria polymorpha. pH effects based modelling predictions of beech earlywood and latewood tissues were also included. Multivariate analysis techniques included response surface optimization, sample-specific standard error of prediction (SEP) method and projection to latent structures partial least squares (PLS) regression. Strong statistical relationships were derived for pH predictions with R² values ranged: from 0.77 to 0.84 for I. hispidus; from 0.77 to 0.84 for T. versicolor and from 0.83 to 0.91 for X. polymorpha. R² values for CIE-based L*a*b* colour space measurements ranged: from 0.43 to 0.69 (L*), 0.66 to 0.76 (a*), 0.42 to 0.53 (b*) for I. hispidus; from 0.59 to 0.69 (L*), 0.69 to 0.79 (a*), 0.64 to 0.79 (b*) for T. versicolor; and from 0.51 to 0.75 (L*), 0.89 to 0.94 (a*), 0.85 to 0.89 (b*) for X. polymorpha. Multivariate technical analysis (response surface analysis, sample-specific SEP, PLS regression) of CIE L*a*b* system and NIRS results should be able to characterize pH effects and surface changes of wood spalted by lignicolous fungi as a quick and reliable non-destructive method relevant to wood-spalting concerns and the forest products industry.     

Differential mixing action effects on functional properties and polymeric protein size distribution of wheat dough

A micro Z-arm mixer and a 2g-Mixograph were used to compare the effect of pin and Z-arm-type mixing actions on mixing properties of wheat flour dough. Although the two mixing curves obtained with pin- and Z-arm-type mixing action showed a very similar mixing trace, no significant correlation was found between the two mixers other than the number of revolutions required for optimal dough consistency (peak resistance). Mixing requirement was described by a rate-independent parameter, the number of revolutions to peak dough development and was found to be greater in a Z-arm mixer than in a pin mixer. Mixing requirement showed significant correlation with stability, which is therefore a dough strength parameter. The change in the polymeric structure of gluten proteins of dough as indicated by %UPP (unextractable polymeric protein percentage) was monitored and showed a smaller decrease with Z-arm mixing than with pin mixing. Therefore, pin-mixing action is more energetic than Z-arm mixing. At peak resistance, Z-arm mixing gives a larger quantity of polymeric protein content in the dough relative to pin mixing. The degree of dough development at maximum resistance in the different mixers was shown to be different. A new parameter, delta-_UPPMZ${\mathrm{UPP}}_{\mathrm{MZ}}$ (the difference between %UPP of dough obtained with pin vs Z-arm mixing actions) was identified and proposed to have some relationship to the stability of the polymeric proteins in the dough.     

陕西关中小麦品种籽粒硬度及测定方法研究

采用四种测定小麦硬度的方法对陕西关中地区33个小麦品种（品系）的籽粒硬度进行了评价，结果表明，陕西关中小麦有9．1属于硬麦，78．8％属于中硬麦，12．1％属于软麦。提出颗粒指数法（PSI），研磨体积法，近红外法（NIR），角质率法测定小麦硬度的评价标准。结果认为：PSI法，研磨体积法和NIR法测定结果的分布对应性较高，NIR法与其它测定方法相关性较高，高角率法有一定局限性。     

Prediction of maize seed attributes using a rapid single kernel near infrared instrument

Non-destructive measurements of seed attributes would significantly enhance breeder selection of seeds with specific traits, and could potentially improve hybrid development. A single kernel near infrared reflectance (NIR) instrument was developed for rapidly predicting maize grain attributes, which would enable plant breeders to quickly select promising individual seeds. With the overall goal being to develop spectrometric calibrations, absorbance spectra from 904 to 1685 nm were collected from 87 maize samples, with 30 kernels of each sample (2610 kernels total), representing a wide variability in the essential amino acids tryptophan and lysine, crude protein, oil and soluble sugar contents. Average sample spectra were matched to bulk reference values. Partial least squares regression (PLSR) calibration models with cross-validation were developed for both relative (% dry matter) and absolute (mg kernel⁻¹) constituent contents. Similarly, models using bagging PLSR were developed. The best model obtained was for relative crude protein content, with an R²p of 0.75 and a SEP of 0.47%. Kernel mass was also highly predictable (R²p=0.76, SEP=0.03 g). Tryptophan, lysine and oil were less predictable, but showed good potential for segregating individual seeds using NIR. Soluble sugar contents produced poor model statistics. Bagging PLSR yielded models with similar levels of prediction.