基于图像熵信息的玉米种子纯度高光谱图像识别

种子纯度是种子质量的一个重要标志,为寻求快速有效的种子纯度识别方法,该文利用高光谱图像技术研究了玉米种子的分类识别问题。首先对17类玉米品种共1632粒种子的高光谱图像提取400～1000nm波长范围内233个波段的熵信息作为分类特征;然后利用偏最小二乘(PLS)投影算法对玉米高光谱图像进行最优波段选择,共获得65个最优波段特征;最后结合偏最小二乘判别分析法(PLSDA)实现了玉米种子的准确识别分类。分类结果表明,在最优波段数仅为全波段27.90%的情况下,其训练精度可以达到99.19%、测试精度为98.90%,可实现多类别样本条件下的玉米种子纯度识别。     

Classification of German white wines with certified brand of origin by multielement quantitation and pattern recognition techniques

A procedure is proposed for the determination of the authenticity of white wines from four German wine-growing regions (Baden, Rheingau, Rheinhessen, and Pfalz) based on their content of some major, trace, and ultratrace elements. One hundred and twenty-seven white wine samples possessing a certificate of origin, all of the 2000 vintage, were analyzed. The concentrations of 13 elements (Li, B, Mg, Ca, V, Mn, Co, Fe, Zn, Rb, Sr, Cs, and Pb) were determined in wine diluted 1:20 by sector field inductively coupled plasma mass spectrometry (SF-ICP-MS). Indium was routinely used as internal standard. Supervised pattern recognition techniques such as discriminant analysis and classification trees were applied for the interpretation of the data. A quadratic discriminant analysis (QDA) allowed the four regions to be discriminated with 83% accuracy when using only eight variables (Li, B, Mg, Fe, Zn, Sr, Cs, and Pb), and the prediction ability for classifying new samples was 76%. By use of a second method, a decision tree, the classification of samples coming from the four regions could be performed with an accuracy of 84% when only four elements were used: Li (very low in samples from Baden), Zn (abnormally low in the samples from the Rheingau), and Mg and Sr (both important for the differentiation between Pfalz and Rheinhessen samples). For this method, the prediction ability was only 74% in the identification of unknown samples. The robustness of the QDA model was not good enough, and therefore the tree is better recommended for the classification of new wine samples from these areas of German wine production.     

Stable isotope and trace element profiling combined with classification models to differentiate geographic growing origin for three fruits: effects of subregion and variety

Classifications of geographic growing origin of three fresh fruits combining elemental profiles with various modeling approaches were determined. Elemental analysis (Ca, Cd, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, P, V, and Zn) of strawberry, blueberry, and pear samples was performed using inductively coupled plasma argon atomic emission spectrometer. Bulk stable carbon and nitrogen isotope analyses in pear were performed using mass spectrometry as an alternative fingerprinting technique. Each fruit, strawberry (Fragaria x ananassa), blueberry (Vaccinium caesariense/corymbosum), and pear (Pyrus communis), was analyzed from two growing regions: Oregon vs Mexico, Chile, and Argentina, respectively. Principal component analysis and canonical discriminant analysis were used for data visualization. The data were modeled using linear discriminant function, quadratic discriminant function, neural network, genetic neural network, and hierarchical tree models with successful classification ranging from 70 to 100% depending on commodity and model. Effects of Oregon subregional and variety classification were investigated with similar success rates.     

调质大米半干法磨粉制备鲜米粉及其品质测定

为考察半干法磨粉对鲜米粉品质的影响,该研究选用旋风磨和布勒磨对调质后含水率为28%和30%的大米进行磨粉,分析大米粉的白度、凝胶特性及糊化特性,对加工鲜米粉的质构特性、蒸煮特性和感官品质进行了分析。结果显示:调质大米可以减小磨粉仪器机械力和热能对大米粉品质的破坏,其中调质大米经布勒磨粉碎后的白度显著高于湿磨粉白度(P0.05);旋风磨含水率为30%调质粉的凝胶硬度最大为3.45 N/cm2,与湿磨粉的无显著差异(P0.05);2种调质粉相比于湿磨粉其崩解值较小,其中布勒磨含水率为30%调质粉回生程度较低,与湿磨粉无显著差异(P0.05)。对于鲜米粉的品质,筛选得到的旋风磨含水率为30%鲜米粉的硬度为35.10 N/cm2,弹性为0.97,较湿磨米粉更柔软弹滑,感官评价总分和蒸煮特性与湿磨米粉无显著差异。以上结果表明,含水率为30%的大米经旋风磨粉碎后能够制得与湿法磨浆相媲美的鲜米粉,可为解决湿磨法废水量大、产品得率低等问题提供参考。     

Corroborative study on maize quality, dry-milling and wet-milling properties of selected maize hybrids

A corroborative study was conducted on the maize quality properties of test weight, pycnometer density, tangential abrasive dehulling device (TADD), time-to-grind on the Stenvert hardness tester (SHT), 100-kernel weight, kernel size distribution, and proximate composition as well as maize dry- and wet-millability by six participating laboratories. Suggested operating procedures were given to compare their measurements and provide the variance structure within and between laboratories and hybrids. Partial correlation coefficient among maize quality properties varied among laboratories. The repeatability and reproducibility precision values were acceptably low for the physical quality tests, except for TADD and SHT time-to-grind measurements. The yields of dry- and wet-milled products and their correlation with maize quality properties were dependent on the collaborating laboratory. This paper highlights the importance of laboratory variation when considering which maize hybrids are best suited for dry-milling and wet-milling.     

基于光谱特性分析的冬油菜苗期田间杂草识别

杂草识别是自动除草的关键环节,运用光谱分析技术可以快速识别杂草。该文以冬油菜苗、冬油菜苗期杂草和土壤为研究对象,通过ASD便携式光谱分析仪采集光谱数据。对每个样本连续采集5组数据,经平均、一阶导数、压缩等预处理后,得到368组波长在400～2300 nm范围内的光谱数据。采用逐步判别分析法,按统计量Wilks’ Lambda最小值原则选择变量,选取了710、755、950和595 nm共4个特征波长。运用4个特征波长分别建立了典型判别函数模型和贝叶斯判别函数模型。用这2组模型分别对预测集进行预测,典型判别函数模型的正确识别率为97.78%,在不同的先验概率下贝叶斯判别函数模型的正确识别率分别为98.89%和97.78%。结果表明：当先验概率根据类别大小计算时,以特征波长建立的贝叶斯判别函数模型能较好的识别冬油菜苗期田间杂草,而且模型稳定。该研究结果可为杂草探测光谱传感器的开发提供参考。     

Considerations regarding the use of hyperspectral imaging data in classifications of food products, exemplified by analysis of maize kernels

Development of robust analytical procedures is critical when using hyperspectral imaging technology in food technology and agriculture. This study used near-isogenic inbred corn lines to address two basic questions: (1) To what extent is classification accuracy increased by grinding maize kernels? (2) Can the classification accuracy of two near-isogenic inbred lines be increased by using a spectral filter to classify only certain hyperspectral profiles from each image cube? Whole kernels and ground kernels in two particle intervals, 0.250-0.354 mm (size 1) and 0.354-0.841 mm (size 2), were examined. Spectral profiles acquired from ground kernels had higher spectral repeatability than data collected from whole kernels. The classification error of discriminant functions from whole kernels was >3 times lower than that of size 1 ground particles. Applying a spectral filter to input data had negligible effect on classifications of hyperspectral profiles from whole kernels and size 2 ground particles, but for size 1 ground particles a considerable increase in accuracy was observed. Independent validation confirmed that distinction between wild type and mutant inbred maize lines could be conducted with >80% accuracy after the proposed spectral filter had been applied to hyperspectral profiles of size 1 ground particles. A combination of discriminant analysis and regression analysis could be used to accurately predict mixture ratios of the two inbred lines. The use of spectral filtering to increase the level of spectral repeatability and the use of hyperspectral imaging technology in large-scale commercial operations are discussed.     

Orange, mandarin, and hybrid classification using multivariate statistics based on carotenoid profiles

A study was undertaken to differentiate citrus on the basis of a carotenoid profile obtained from the HPLC determination of 12 carotenoids found in saponified fresh juice. Mandarin, orange, and various hybrid varieties were analyzed to determine their carotenoid profiles. The resulting peak areas were analyzed using principal component analysis (PCA), canonical discriminate analysis (CDA), and Mahalanobis distances. These were used to develop models for classifying the juices into appropriate groups. Thirty-two samples were analyzed to determine classification techniques. Mandarin and orange juices were quite distinct, with the hybrids scattered throughout the mandarin and orange clusters using PCA. CDA resulted in three distinct groups with only a few of the hybrids in the orange grouping.     

我国北方37个高产春玉米品种干物质生产及氮素利用特性

选育氮高效品种是实现玉米高产高效生产的根本途径。为探明我国北方目前主推高产春玉米的物质生产及氮素利用特性,本研究选择该区域高产品种37个,采用盆栽试验,依粒重和氮素子粒生产效率划分其类型,分为高产高效（I）、高产中效（Ⅱ）、中产中效（Ⅲ）及低产低效（Ⅳ）4种类型。其中,中产中效型品种最多,为56.8%;高产高效型品种最少,仅为8.1%;高产中效型和低产低效型品种分别为13.5%和21.6%。4个类型品种干物质生产及氮素利用效率开花前差异不显著,开花后是产生差异的关键时期;成熟期I型品种干物质和氮向子粒的分配比例较高,而Ⅳ型品种向根和茎秆的分配比例较高。同时,I型品种的氮转移量、氮转移效率和贡献率显著高于其他3类型品种。经相关和通径分析,氮素干物质生产效率、粒重及氮含量与氮素子粒生产效率显著相关。所以,较高的粒重和较低的植株氮含量是高产氮高效品种的基本特征。     

Maize ear rot and moniliformin contamination by cryptic species of Fusarium subglutinans

Fusarium subglutinans causes maize ear rot and contaminates grain with the mycotoxin moniliformin. Previous DNA sequence analysis divided F. subglutinans from maize into two cryptic species, designated groups 1 and 2. Here, it was determined whether the two groups differ in the agriculturally important traits of virulence on maize and moniliformin production in planta. Thirty-seven strains from U.S. maize were assigned to groups 1 and 2 by DNA sequence analysis. In field tests, all strains were highly virulent on maize inbred B73 and four maize hybrids. In planta, 82% of group 1 strains and 25% of group 2 strains produced high levels (100-1500 microg/g) of moniliformin. All group 2 strains from more northern states produced little or no moniliformin (0-5 microg/g). These data indicate that moniliformin production is highly variable in F. subglutinans from U.S. maize and that production may not be required for the fungus to cause maize ear rot.     

Volatile constituents in fresh and processed juices from grapefruit and new grapefruit hybrids

Forty-five volatile constituents of juices from grapefruit and grapefruit hybrids were quantified by headspace gas chromatography. The three types of grapefruit juice analyzed include pasteurized juice not from concentrate, reconstituted single strength juice from concentrate, and fresh, unpasteurized juice. Principal component and discriminant analyses were carried out using 48 grapefruit juice samples, and the samples were classified into the three types of juice based on degree of processing. Discriminant analysis was superior to principal component analysis for this purpose. Juices from two recently developed grapefruit hybrids were classified similarly to unpasteurized grapefruit juices from commercial cultivars.     

Authentication of Rice by Three-Sided Image Analysis of Kernels Using Two Mirrors

The authentication of rice by imaging three sides (top, side, and front) of a rice kernel using two mirrors was attempted. Nine image characteristics (area, aspect ratio, minimum diameter, maximum diameter, perimeter, and red, green, and blue values) for each side of a rice kernel were measured. Japonica type rices were used for the study; a total of 89 rice samples were used: 40 samples harvested in Korea, 33 in America, 5 in Australia, and 11 in China. Image data collected from 105 kernels per rice sample were used as a calibration data set to develop a model for quadratic discriminant analysis. In overall pairwise trials, on average, 85.8–94.5% correct identification of the cultivated country was obtained, depending on the number of image characteristics used for the discriminant model. For the effective discrimination of the cultivated country, 13 image characteristics were selected. As a test sample set, 20 kernels per sample were used. A model for discriminant analysis with 13 image characteristics yielded an average of 95.0 and 97.5% accuracy for calibration and test sample set, respectively.     

玉米杂交种氮效率基因型差异

利用北京地区推广的 8个主要玉米杂交种 ,在不施N和施N条件下研究了产量、N累积量及N效率的基因型差异。结果表明 ,在不施N和施N条件下 ,不同品种之间产量的变异幅度分别为 62%、57% ;N累积总量的变异幅度则分别为 55%、48%。其中以农大 108的产量和N累积总量最高 ,中单 120和中单 2号最低。随施N量增加 ,吸收效率和N效率显著下降 ,利用效率则变化不大。在 2个施N水平下 ,N效率基因型差异主要决定于吸收效率     

Collection, evaluation and classification of Greek populations of faba bean (Vicia faba L.)

Fifty-five Greek Vicia faba L. populations, collected from diverse areas, were planted at two dry and low fertility sites for evaluation and classification. Yield evaluation, which was carried out by Principal Component Analysis (PCA) on the basis of seven yield traits, showed the number of pods per plant, number of ovules and seeds per pod, and branching from the basal nodes to be the most important traits for population evaluation regarding yield. For population classification, four dissimilarity coefficients (Manhattan, Average Taxonomic Distance, Euclidean distance and squared Euclidean distance) and four multivariate methods (PCA, UPGMA, Neighbor-joining and Principal Coordinate Analysis) were evaluated using fifteen morphological and seven yield traits. Neighbor-joining was chosen as the most suitable multivariate method. This method combined with PCA for the seven yield traits, placed the populations into six groups. As revealed by the application of PCA on all twenty-two traits the grouping was based mainly on pod characteristics, stem thickness, plant height, 1000 seed weight and branching from basal nodes. Based on the results of the present study, a model is proposed for conserving cross-pollinated species, such as faba bean.     

Event-specific quantitative detection of nine genetically modified maizes using one novel standard reference molecule

With the development of genetically modified organism (GMO) detection techniques, the Polymerase Chain Reaction (PCR) technique has been the mainstay for GMO detection, and real-time PCR is the most effective and important method for GMO quantification. An event-specific detection strategy based on the unique and specific integration junction sequences between the host plant genome DNA and the integrated gene is being developed for its high specificity. This study establishes the event-specific detection methods for TC1507 and CBH351 maizes. In addition, the event-specific TaqMan real-time PCR detection methods for another seven GM maize events (Bt11, Bt176, GA21, MON810, MON863, NK603, and T25) were systematically optimized and developed. In these PCR assays, the fluorescent quencher, TAMRA, was dyed on the T-base of the probe at the internal position to improve the intensity of the fluorescent signal. To overcome the difficulties in obtaining the certified reference materials of these GM maizes, one novel standard reference molecule containing all nine specific integration junction sequences of these GM maizes and the maize endogenous reference gene, zSSIIb, was constructed and used for quantitative analysis. The limits of detection of these methods were 20 copies for these different GM maizes, the limits of quantitation were about 20 copies, and the dynamic ranges for quantification were from 0.05 to 100% in 100 ng of DNA template. Furthermore, nine groups of the mixed maize samples of these nine GM maize events were quantitatively analyzed to evaluate the accuracy and precision. The accuracy expressed as bias varied from 0.67 to 28.00% for the nine tested groups of GM maize samples, and the precision expressed as relative standard deviations was from 0.83 to 26.20%. All of these indicated that the established event-specific real-time PCR detection systems and the reference molecule in this study are suitable for the identification and quantification of these GM maizes.     

Use of chemical profiling to differentiate geographic growing origin of raw pistachios

The objective of this study was to demonstrate the feasibility of chemical profiling methods combined with multivariate methods to differentiate the geographical growing regions of pistachios (Pistachia vera). Elemental analysis (Ba, Be, Ca, Cu, Cr, K, Mg, Mn, Na, V, Fe, Co, Ni, Cu, Zn, Sr, Ti, Cd, and P) of pistachios samples was performed using inductively coupled plasma atomic emission spectrometry. Analysis of inorganic anions and organic acids (selenite, bromate, fumarate, malate, selenate, pyruvate, acetate, phosphate, and ascorbate) of pistachio samples was performed using capillary electrophoresis. Bulk carbon and nitrogen isotope ratios were performed using stable isotope MS. There were nearly 400 pistachio samples analyzed from the three major pistachio growing regions: Turkey, Iran, and California (United States). A computational evaluation of the trace element data sets was carried out using statistical pattern recognition methods including principal component analysis, canonical discriminant analysis, discriminant analysis, and neural network modeling. Several linear discriminant function models classified the data sets with 95% or higher accuracy. We report the development of a method combining elemental analysis and classification techniques that may be widely applied to the determination of the geographical origin of foods.     

基于数据挖掘分类法的农用地分等

应用决策树模型、BP神经网络和Logistic回归模型等分类法,对龙川县农用地分等进行了实证研究,并对各方法的分等结果有效性进行了评价,同时利用混淆矩阵探讨了样本数量对3种模型分类精度的影响。结果表明,样本数量对模型影响有差异,其中对BP神经网络和决策树模型影响较大,在较多训练样本时,模型的精度较高。在较多样本支持下,BP神经网络精度最高,但训练模型的时间较长,可解释性差;决策树模型既具有较高的精度又具有良好的可解释性;Logistic回归模型表现较差。决策树模型最适合龙川县农用地分等工作。研究结果表明,数据挖掘分类法是有效而准确的土地评价方法,有助于提高土地评价的精度和准确性,对农用地分等方法的优化具有一定的借鉴意义。     

Characterization of protein fractions from Bt-transgenic and non-transgenic maize varieties using perfusion and monolithic RP-HPLC. Maize differentiation by multivariate analysis

Protein fractions from transgenic Bt and non-transgenic maize varieties, extracted by the Osborne solvent fraction procedure, were characterized for the first time by perfusion and monolithic RP-HPLC in very short analysis times. Albumins and globulins from different transgenic Bt maizes as well as from their non-transgenic isogenic varieties were eluted in four peaks using perfusion RP-HPLC, whereas prolamins and glutelins were separated in seven peaks. Monolithic RP-HPLC enabled the separation of maize proteins in a large number of peaks showing 6 and 10 main peaks for albumins and globulins, respectively. Prolamins migrated at retention times higher than 5 min as seven peaks, whereas glutelins were separated in three main peaks appearing at retention times higher than 6.0 min. Moreover, chromatograms of the whole protein extracts showed 8 and 11 components for perfusion and monolithic RP-HPLC, respectively. A comparison of the chromatograms of the whole protein extracts relative to transgenic and non-transgenic varieties evidenced quantitative differences on the percentages of area, mainly for peaks 2 and 3 by perfusion RP-HPLC and for peaks 3 and 7 by monolithic RP-HPLC. A discriminant analysis based on these proteic profiles was carried out to classify and predict transgenic Bt maize lines, achieving 100% correct classification using perfusion RP-HPLC.     

云南杂交玉米SSR标记核心引物的筛选及多样性分析

为降低合成引物成本,加快杂交玉米品种鉴定进程,筛选适用于云南玉米杂交种SSR标记鉴定的核心引物。本研究用玉米SSR标记鉴定规程中推荐的40对SSR引物对12个杂交玉米品种DNA进行扩增,筛选出多态性高、扩增效果好、稳定性好的20对引物,推荐其作为云南玉米杂交种特异性鉴定的核心引物。并用筛选的核心引物对云南220个杂交玉米材料进行多样性分析及评价。筛选的20对SSR引物在12个杂交玉米品种中共检测到216个等位变异,平均每对引物检测到10.80个。20对SSR核心引物在220个杂交玉米材料中共检测到236个等位变异,平均每对引物检测到11.80个;平均有效等位变异为3.158 8;平均Shannon-Weaver多样性信息指数为3.028 2。聚类分析表明,220个杂交玉米材料间的相似系数为0.62~0.95。以0.62为阈值可将220个杂交玉米材料划分为两大类群,第Ⅰ类群包括10个玉米材料,占4.5%;第Ⅱ类群包括210个玉米材料,占95.5%。本研究筛选的20对SSR引物多态性高、扩增效果好、稳定性好,可作为云南玉米杂交种特异性鉴定的核心引物。