基于近红外高光谱成像及信息融合的小麦品种分类研究

高光谱成像技术因具有图谱合一的特点在作物品种鉴别方面具有较大潜力,但目前研究大多只提取利用了光谱信息,对图像信息没有进行有效利用。本文利用近红外高光谱成像仪采集了强筋、中筋、弱筋3个类型共计6个品种的单粒小麦种子高光谱图像,提取了长、宽、矩形度、圆形度、离心率等12个形态特征,并对图像中的胚乳和胚区域进行分割建立掩膜,提取了胚乳和胚区域的平均光谱信息。采用PLSDA和LSSVM方法建立基于图像信息的判别模型,结果表明强筋、弱筋两者二分类的识别率能达到98%以上,强筋、中筋两者二分类的识别率只能达到74.22%,说明近红外高光谱图像的形态信息能够反映品种间差异,但单独利用图像信息进行分类时准确度可能欠佳。采用SIMCA,PLSDA和LSSVM方法建立了胚乳和胚区域光谱信息的多分类模型,胚乳区域的分类效果较胚区域略好,说明籽粒不同部位的形状差异会影响分类效果。进一步融合光谱信息和图像信息,采用SIMCA,PLSDA和LSSVM方法建立融合模型,识别率较单独的图像或光谱信息模型均略有提升,PLSDA方法从原来的96.67%提升到98.89%, 表明充分挖掘高光谱图像所包含的形态特征和光谱特征可有效提高分类效果。     

时间分辨傅氏变换红外光谱方法中的定时直流成分

在常规的傅里叶变换红外光谱学中，干涉图的直流成分为∫＾∞０Ｂ（σ）ｄσ，而在时间分辨ＦＴＩＲ中，直流成分Ｉｄｅ成为∫（∞，０）Ｂ（σ，τ）ｄσ，即与信号出现的时间τ有关，称为定量直流成分。     

一种全场三维共焦检测的新方法

提出了一种利用微光学阵更合成器件实现全场三维面形并行检测、记录的共焦方法。此方法的关键是，在共焦系统中引入一一微光学耻列合成器件来产生一点光源阵列，实现了对同一一剖面同时并行检测，并首次采用了ＣＣＤ面阵像元取代上孔光阑直接截取三维信息光强。本文讨论了该方法的三信息检测原理，响应关系，给出了初步的实验结果和三维重构图。     

Photoelectron diffraction study of a catalytically active overlayer: C2H2 on Pd{111}

A quantitative structure determination of a newly discovered (2×2) adsorption phase of acetylene chemisorbed on Pd{111} has been performed by scanned-energy mode photoelectron diffraction: this phase corresponds to the threshold coverage for the catalytic conversion of acetylene to benzene. The carbon atoms in the C₂H₂ molecule are located almost over bridge sites with a C–C bond length of 1.34+0.10 Å, the centre of the molecule being positioned almost over a hollow site. Of the two hollow sites the hcp site (directly above a second layer Pd atom) is favoured, particularly by a subset of the data most sensitive to this aspect of the structure, but the full analysis indicates that the fcc site (above a third layer Pd atom) cannot formally be excluded. The adsorption site adopted by acetylene in the higher coverage phase on Pd{111} is essentially identical. This is the dominant structure in the coverage regime which is catalytically active for the conversion of acetylene to benzene. The implications of these findings for acetylene coupling reactions over Pd{111} are discussed.     

Diffusional coagulation of superparamagnetic particles in the presence of an external magnetic field

In this paper, we investigate the diffusional coagulation of colloidal superparamagnetic (SP) latex particles that are under the influence of an external magnetic field. The cluster size distributions (CSDs) that evolve with time were determined using an optical set-up that permits the direct visualization of particle clusters. Following the dynamic scaling analysis of van Dongen and Ernst (Phys. Rev. Lett. 54 (1985) 1396), we find that the CSDs all collapse onto a master curve when properly scaled. The bell-shape of this master curve indicates that large clusters preferentially scavenge small clusters in our system. From the time evolution of the average cluster size we infer that the reactivity between large clusters diminishes with increasing cluster size. These results are consistent with a simple mathematical formulation of the coagulation rate constant, or kernel, for the Brownian coagulation of magnetic particles. Moreover, our results support a growing body of evidence that the dynamic scaling theory developed by van Dongen and Ernst is a useful framework with which to study the microscale processes governing particle coagulation.     

中子引起单粒子翻转过程的Monte Carlo计算模拟

从中子与硅原子相互作用的物理机理出发,利用Monte Carlo方法编制了中子引起单粒子翻转的计算模拟程序,并对14 MeV中子环境下的16K位静态存储器硅片翻转过程中的物理量进行了计算,同时可为中子引起的单粒子翻转的研究提供截面和描述内部物理过程的参考数据。     

红外光谱在纤维质文物材料鉴别中的应用研究

纺织纤维和纸张纤维是常见纤维质文物材料, 是构成博物馆精美文物如服饰手稿书画的基本材料, 近年来寻求通过无损或微损方法对这一类材料的鉴别以及劣化状况评价备受文物鉴赏家和文物保护工作者的关注。借助傅里叶变换红外光谱, 研究博物馆常见纺织纤维材料棉、麻、桑蚕丝、柞蚕丝、羊毛的红外光谱特征和它们的分子结构组成异同, 研究传统纸纤维稻草、麦草、龙须草、龙旗松、桑皮红外光谱特征。结果表明: 衰减全反射傅里叶变换红外光谱无损分析技术可通过比较3 300～2 800 cm-1 CH, NH, OH振动区间光谱形状以及指纹区峰位以区别不同种类纺织品纤维;碳氧振动纸张纤维最明显光谱差异位置出现在与纤维素OH伸缩振动相关波数3 300 cm-1和与C—O—C相关波数1 332, 1 203, 1 050 cm-1。文章探索研究红外光谱技术结合主成分分析法在快速鉴别纤维材料中的应用。通过对全光谱数据多元散射校正(MSC)预处理后进行主成分分析, 可以把红外光谱十分相似的纺织纤维棉和亚麻、桑蚕丝和柞蚕丝明显分类;对光谱相似的纸纤维, 可采用选择不同光谱波数段进行主成分分析, 比较发现能够把五种纸纤维明显区分的光谱区间为3 800～2 800 cm-1。本研究为分子光谱无损分析技术应用于文物材料鉴别、科学评估纤维材料保存状况提供基础研究。     

基于动态磁聚焦的CT球管飞焦点研究

作为CT设备的核心器件,CT球管采用动态磁聚焦技术以利于大电流输出时小焦点的实现。飞焦点技术可以多角度记录每次扫描的每个投影,提高采样率,从而大幅改善图像清晰度,提高成像质量。当CT球管为1 A大电流输出时,采用CST软件对双磁四极透镜关键参数进行模拟仿真及优化,满足焦点目标尺寸为0.2 mm×0.6 mm的同时具备飞焦点功能。     

基于太赫兹时域光谱技术的煤岩识别方法研究

煤岩识别一直是制约煤矿无人化开采的关键问题之一。传统的人工采煤因为工作环境极其复杂,很难精准地找到煤岩的分界面,容易造成欠切割或过切割现象。太赫兹光谱技术作为一种无损探测技术,能够反映出被测物体的物理和化学信息,可以成为研究煤岩识别的有效方法。采用太赫兹时域光谱技术与多元统计法—聚类分析（CA）和主成分分析（PCA）相结合的方法来识别不同种类的煤岩。通过透射式太赫兹光谱仪获得六种煤岩样品的太赫兹光谱,对其进行FFT等一系列数学计算可以得到各种样品的折射率、吸收系数以及介电常数。计算结果表明不同种类的煤岩在折射率、吸收系数上都存在差异。分析各类煤炭样品的折射率和吸收系数与样品的各组成成分含量之间的关系,可以发现碳含量是影响其样品折射率大小的因素之一,灰分含量是影响其样品吸收系数大小的因素之一。聚类分析中两类样品的欧氏距离与主成分分析中的第一主成分（PC1）得分都能反映煤岩样品之间的相似性和相异性,并且CA与PCA的结果保持一致。分别将各类样品在0.5～2.5 THz频率范围内的折射率、吸收系数与CA和PCA结合,组成太赫兹数据与煤岩之间的模型。分析表明：根据不同样品之间的相似性,两种模型中六种煤岩样品均被分为两类;在各种样品的吸收系数与CA-PCA组成的模型中,四种煤炭被聚集在一起,并且石英砂岩（GSR-4）具有很好的独特性：石英砂岩拥有最小的PC1得分值以及石英砂岩与第二类之间的欧氏距离最大,为219.03。由此可见采用太赫兹技术与多元统计方法结合,可以实现煤岩的准确识别,识别准确率可以达到100%。     

New ultrasonic assisted technology of freezing,cooling and thawing in solid food processing: A review

Solid foods include fish, shrimp, shellfish, and other aquatic products, fruits, and vegetables. These products are commonly used for food freezing, cooling, and thawing. However, traditional freezing, cooling, and thawing of solid food technologies have limitations in quality, such as protein denaturation and water loss in food. Ultrasound-assisted technology has become a useful method in solid food processing due to improved preservation quality of solid food. This paper comprehensively reviews the mechanism and application of ultrasonic in solid food processing technology. Although the application of ultrasound-assisted ultrasound in solid food processing is relatively comprehensive, the energy saving of food cold processing is essential for practical application. This paper analyzes the optimization of ultrasonic in solid food processing, including orthogonal/multi-frequency technology and the combination of ultrasonic and other technologies, which provides new ideas for freezing, cooling, and thawing of solid food processing.