贝类腹泻性毒素现场高通量前处理及快速检测仪器设计

为实现贝类腹泻性毒素的现场快速检测,本论文设计了一种现场高通量前处理装置及基于移动终端的快速检测系统。通过使用酶联免疫反应技术对腹泻性贝毒进行显色反应,结合基于移动终端的快速检测装置,采集显色反应结果图像,实现准确快速的检测结果分析。采用所设计的前处理装置与实验室手工前处理标准方法对比,其回收率分别为89％和93％,结果表明,本文设计的前处理装置能够满足现场检测分析的需求。通过加标样品检测,采用移动终端的快速检测系统测量标准差为0.13,平均回收率89.5％。实验结果表明,该检测系统的准确性和重复性满足实际检测的需求。最后,针对实际样品,与酶标仪检测结果相比较,表明了所设计的现场高通量前处理装置及快速检测系统能达到对贝类腹泻性毒素现场快速精确检测,为贝类腹泻性毒素现场检测提供了新的方法和仪器。     

基于移动终端的麻痹性贝类毒素现场快速检测系统设计

为实现食品中毒素的现场快速检测,设计了一种基于移动终端的麻痹性贝类毒素现场高通量快速检测系统.采用PSP试剂盒对毒素进行显色反应,实现了一定浓度范围内的毒素检测;采用暗室结构,配合面光源透射,提供了稳定的检测环境;采用具有较高像素摄像头的移动终端对反应结果进行采集和处理,实现了检测结果的精确分析和高效交互.以实际样品作为对象进行了系统的功能测试,并与商用酶标仪的结果进行了对比,证明该系统能够较为准确地测得样品中麻痹性毒素的浓度,且检测速度较快,灵敏度较高.结果表明,本系统操作简单,且具有较高的灵敏度、检测速度和通量,能在现场实现麻痹性贝类毒素的快速检测,为食品安全评估提供了一种高效的初筛手段.     

基于上转换发光技术的便携式食品安全检测仪设计

针对当前食品安全快速检测设备存在灵敏度低、无法准确定量检测等问题,设计了一种定量、快速的便携式食品安全检测仪,其能够快速筛查食品中多种食品危害因子。利用上转换材料处理食品样品产生特征荧光,设计了以STM32F103C8T6单片机为核心的检测仪进行检测荧光信号,包括信号检测光路系统和信号处理电路系统,以及嵌入式主控程序软件设计。实验结果表明:检测仪功耗低,体积小,准确度高,实际样品检测回收率在97. 5%～105%之间,满足快速检测的需求。     

一种针对手写体图像的取证算法

数字图像取证是信息安全领域的研究热点.在手写体图像方面,现有的取证算法对于图像采用各种重采样技术的篡改,其检测效果并不理想.在本文中,我们根据源区域和篡改区域的特征值不变性,提出一种简单有效的盲取证算法.该算法实现了手写体图像篡改的自动检测和篡改区域的定位,并且比现有的重采样检测方法对相同的手写体图像具有更好的检测率,尤其是在图像中的字有模糊、缺损的情况下检测的优势更明显.实验结果表明,对于经过各种重采样处理的手写体图像,该算法比现有取证算法的检测率高20%,当虚警率小于1%时,本算法的检测率达96.9%以上.     

基于微变电容式传感器的角度-振动测量方法

利用微变电容式MEMS传感器,设计了一种角度-振动量的检测方法。以RC串联回路直流充放电理论为基础,单片机系统为主体,实现振动量和角度偏移量的检测。经过适当的滤波算法对原始信号进行处理,可以有效地减小噪声的干扰,并快速而准确地分离出基波分量和高次谐波分量,进而得出较好的振动曲线和角度曲线。经实验验证：角度检测和振动检测可同时进行且互不干扰。     

基于局部堆内存抽象表示的堆操作程序内存泄露检测

堆操作程序通过共享易变数据结构可灵活地申请、合并、删除堆内存.这类程序的内存泄漏检测要求精确的域敏感的指针别名信息,变得尤其复杂和难以处理.针对这个问题,提出了基于"指针扩展类型"域敏感的堆内存抽象方法,对指针变量在形态上的排列关系进行抽象以支持堆的局部推理.首先,定义了各种基本语句的操作语义,然后基于该抽象方法采用前向数据流迭代算法提出了一种新的内存泄露检测算法.在Crystal编译框架下实现了面向C程序的内存泄漏检测原型工具Heapcheck,该工具支持复杂数据结构内指针型数据域上的内存泄露检测.在典型基准C程序上的实验结果分析表明,该方法与现有的技术相比在效率和精度上都具有优势.     

安捷伦科技发布全面的样品前处理参考指南

正安捷伦科技推出由样品前处理专家Ronald E.Majors博士撰写的《色谱样品前处理的基本知识(Sample Preparation Fundamentals for Chromatography)》。该本著作是色谱样品前处理方法的全面参考指南。Majors博士表示:"这本指南介绍当今一些最常用的样品前处理技术。该指南展示了从‘稀释和上样’到多级样品处理这类技术的正确使用方法,以便获得最佳结果。"本指南着重介绍了样品前处理的基本信息,以及有关     

认知推理的家庭服务机器人演示学习研究

提出一种基于对比示范的演示学习方法,旨在指导家庭服务机器人快速智能地学习指定任务决策.在模拟家庭环境下,首先利用视觉手段检测和跟踪任务对象,获取”教师”演示正反例关键信息,通过动作识别和状态检测将信息抽象成符号形式,然后经过认知推理得到演示任务的执行决策和约束条件,并通过动作共享指导机器人重现任务.给定演示实例,在真实移动机器人平台上进行了多次成功演示验证,结果表明本方法在家庭服务机器人快速准确的任务学习方面是有效的.     

实心车轴的超声相控阵探伤

针对机车车轴的缺陷存在区域及其分布特点,分析利用超声相控阵技术进行机车车轴探伤的方法,提出一种新的机车车轴的相控阵超声检测方案,仅采用一个相控阵探头即可完成整个车轴的缺陷探测,且定位准确.结果表明,该系统可一次性完成对样板轴的全扫查,得出各缺陷所在位置及深度等定位信息,根据车轴缺陷检测数据结果可快速在车轴上找到缺陷,检测效率和检测精度较高,满足车轴检测工艺要求.     

基于脉冲耦合神经网络的工件边缘定位

为实现在工业测量中对工件边缘的快速准确定位,以满足在线测量系统的实时性的要求,利用脉冲耦合神经网络(PCNN,Pulse Coupled Neural Network).并结合前三阶灰度矩实现了对工件边缘的精确定位.该方法首先利用脉冲耦合神经网络对待测量工件进行初步定位,然后利用初步定位的结果设置感兴趣的区域,再利用前三阶灰度矩在原始灰度图像上的感兴趣区域内进行边缘的亚像素细分.实验结果表明该方法抗噪声能力强,边缘定位准确,同时能够提高边缘的检测速度.     

A free space frequency‐time‐domain technique for electromagnetic characterization of materials using reflection‐based measurement

In this report, a free space frequency‐time‐domain technique is presented for characterizing the electrical properties and thickness of the sample using multiple reflections and fabry‐perot resonance phenomenon. The retrieval of constitutive electromagnetic parameters of the sample has been carried out by comparing the measured reflection coefficient data from the sample at two different incident angles. The relative permittivity as well as relative permeability along with the thickness of different samples viz., beryllia, silicon, and plexiglass have been evaluated with high accuracy in the frequency range 1 to 15 GHz. The method is also experimentally validated by successfully reconstructing the unknown material properties of two different samples. The unique advantage of this method lies in non‐requirement of any prior knowledge of the sample's thickness for measuring the complex relative dielectric constant as well as relative permeability of the sample. To determine the electromagnetic properties of the sample, the sole knowledge of reflection coefficient data are needed. Moreover, the method does not involve any additional measurement for the reference calibration. The simple, cost‐effective proposed scheme is quite useful in many applications like accurate determination of signal strength in indoor wireless communication, through wall imaging, food industry, and so on.     

Characterization and analytical validation of a microcantilever-based sensor for the determination of total carbonate in soil samples

A microcantilever device has been used as a miniaturized sensor for the determination of total carbonate in soil samples. The method is based on the selective generation of CO₂ (g) from samples and the measure of the pressure effect on the microcantilever sensor located in a closed system. The experimental set-up uses a manifold including the sample reaction minichamber and a dedicated flow-cell designed and built for housing the microcantilever. The response of this sensor was electronically and analytically characterized by using conductance signals. The analytical performance of the sensor was compared to that provided by the conventional quartz crystal microbalance (QCM), demonstrating advantages in terms of sensitivity, linear range of response and resolution. After the optimization of the method, it was validated for its use for direct determination of carbonate in soil samples. The method allowed the determination of carbonate in the 3–75 mg range, with a precision of 1.7%, expressed as relative standard deviation. It was applied to the analysis of different soil samples, obtaining results in agreement with those produced by the official method.     

基于PCA的人脸识别系统的研究与实现

人脸识别技术是人脸识别系统以人脸图像作为识别身份的媒介。研究了PCA算法的原理,基于matlab平台实现了人脸识别系统。该系统在对图像进行预处理的基础上,以ORL人脸库中部分图像为样本,采用K-L变换计算训练样本特征值及特征向量,得到特征脸向量,进而计算测试样本投影到特征脸空间的坐标系数,最后分类识别得出结果。实验证明该系统识别率较高,对实际应用有一定的参考价值。     

Application of bootstrap method in conservative estimation of reliability with limited samples

Accurate estimation of reliability of a system is a challenging task when only limited samples are available. This paper presents the use of the bootstrap method to safely estimate the reliability with the objective of obtaining a conservative but not overly conservative estimate. The performance of the bootstrap method is compared with alternative conservative estimation methods, based on biasing the distribution of system response. The relationship between accuracy and conservativeness of the estimates is explored for normal and lognormal distributions. In particular, detailed results are presented for the case when the goal has a 95% likelihood to be conservative. The bootstrap approach is found to be more accurate for this level of conservativeness. We explore the influence of sample size and target probability of failure on the quality of estimates, and show that for a given level of conservativeness, small sample sizes and low probabilities of failure can lead to a high likelihood of large overestimation. However, this likelihood can be reduced by increasing the sample size. Finally, the conservative approach is applied to the reliability-based optimization of a composite panel under thermal loading.     

X射线衍射外推法精确测定晶胞参数

本文提出的晶胞参数精确测定方法是用测定晶体各个衍射谱线，利用新的外推函数消除系统误差，并根据最小二法求得最佳外推直线，得到精确的晶胞参数，实验分析表明，用这种方法可以得到精确的晶胞参数。     

双目系统全视域采样的SVM标定模型分析

目的 为改善摄像机间接标定采样不全、模型表达模糊问题,实现小视场下检测视域完备采样,提出一种基于双目系统全视域采样的支持向量机（SVM）标定方法。方法 该方法利用六角晶格标定板靶点序号可读特点为基础,采集整个双目系统有效视域中检测点的视差坐标、世界坐标并建立完备的样本集。选取SVM对该样本集进行训练,将SVM算法得到的模型参数代入其决策函数中进行求解,获得公式化的标定模型。由于六角晶格标定板的四角和中心分布了5个互为非中心对称的多边形,可在标定板部分区域被采集的情况下获取标定板位姿信息,进而读取采集的各靶点序号。通过上下移动标定板,利用HALCON算子获取图像中各靶点的序号,建立双目视觉系统检测区域的完备样本集。最后,利用SVM算法训练样本获得标定模型,可以明确表达出标定模型的数学形式。结果 与传统采样建立的模型进行对比分析,实验结果表明该方法建立模型的标定误差减小了24.51%,降低了标定模型在传统方法未采样区域的标定误差,证明了该方法的可行性。结论 提出一种基于双目系统全视域采样的支持向量机标定方法,通过非中心对称的多边形确定标定板上靶点的序号,实现双目视觉系统检测视域的完备采样。实验结果表明该方法提高了摄像机间接标定的精度,具有良好的适用性和鲁棒性,适用于小视域内双目视觉系统的间接标定。     

Photoacoustic determination of tension in magnetic tape

A non-contact method, using photoacoustically generated antisymmetric Lamb waves detected with a laser Doppler technique, is used for determining tension in magnetic tape. For the tape samples used in this study, the bending stiffness contributes as much to the overall stiffness as a 1 Newton tension force, so an accurate determination of the tension must also necessarily include an independent measurement of the bending stiffness. Our data shows that the flexural velocity normal to the direction of the applied tension provides a determination of the bending stiffness independent of tension. The dependence of the wavespeed on tension is measured, showing the expected tension dependence and demonstrating the utility of this approach.     

A plug-based microfluidic system for dispensing lipidic cubic phase (LCP) material validated by crystallizing membrane proteins in lipidic mesophases

This article presents a plug-based microfluidic system to dispense nanoliter-volume plugs of lipidic cubic phase (LCP) material and subsequently merge the LCP plugs with aqueous plugs. This system was validated by crystallizing membrane proteins in lipidic mesophases, including LCP. This system allows for accurate dispensing of LCP material in nanoliter volumes, prevents inadvertent phase transitions that may occur due to dehydration by enclosing LCP in plugs, and is compatible with the traditional method of forming LCP material using a membrane protein sample, as shown by the successful crystallization of bacteriorhodopsin from Halobacterium salinarum. Conditions for the formation of LCP plugs were characterized and presented in a phase diagram. This system was also implemented using two different methods of introducing the membrane protein: (1) the traditional method of generating the LCP material using a membrane protein sample and (2) post LCP-formation incorporation (PLI), which involves making LCP material without protein, adding the membrane protein sample externally to the LCP material, and allowing the protein to diffuse into the LCP material or into other lipidic mesophases that may result from phase transitions. Crystals of bacterial photosynthetic reaction centers from Rhodobacter sphaeroides and Blastochloris viridis were obtained using PLI. The plug-based, LCP-assisted microfluidic system, combined with the PLI method for introducing membrane protein into LCP, should be useful for minimizing consumption of samples and broadening the screening of parameter space in membrane protein crystallization.     

面向深度学习的对抗样本差异性检测方法

深度神经网络（DNN）在许多深度学习关键系统如人脸识别、智能驾驶中被证明容易受到对抗样本攻击,而对多种类对抗样本的检测还存在着检测不充分以及检测效率低的问题,为此,提出一种面向深度学习模型的对抗样本差异性检测方法。首先,构建工业化生产中常用的残差神经网络模型作为对抗样本生成与检测系统的模型;然后,利用多种对抗攻击攻击深度学习模型以产生对抗样本组;最终,构建样本差异性检测系统,包含置信度检测、感知度检测及抗干扰度检测三个子检测系统共7项检测方法。在MNIST与Cifar-10数据集上的实验结果表明,属于不同对抗攻击的对抗样本在置信度、感知度、抗干扰度等各项性能检测上存在明显差异,如感知度各项指标优异的对抗样本在置信度以及抗干扰度的检测中,相较于其他类的对抗样本表现出明显不足;同时,证明了在两个数据集上呈现出差异的一致性。通过运用该检测方法,能有效提升模型对对抗样本检测的全面性与多样性。