Processing of thermogram sequences to determine the rotation rate of a vortex flame

We study vortex flame radiation in laboratory conditions. A fire torch was burning from a tank, which was fixed at the axis of uprising swirl airflow. Ethanol was used as a fuel. Processing of a sequence of thermograms of a swirling torch is described. The vortex flame revolutions were calculated using FFT of time pulsations of a signal from a thermal imaging system.     

一种基于HSI颜色模型的目标提取方法

在工业、农业、军事以及医疗等领域中,目标识别是一个关键技术,近年来在图像处理方面已成为一个重要的研究课题。提出了一种用于除草机器人的基于HSI(色调、明度和饱和度)颜色模型的彩色图像目标提取方法。该方法首先将具有某种色彩信息的杂草与农作物从背景土壤中分离出来,然后再根据杂草与农作物形状的不同,采用腐蚀和膨胀的方法进一步将杂草与农作物分离。该方法除了用于除草机器人以外,还可用于水果采摘等。仿真结果验证了该方法的有效性。     

一种用于光纤链路振动信号模式识别的规整化复合特征提取方法

相位光时域反射链路监测系统是一种利用光纤作为传感介质的传感系统, 能够监测、定位、识别入侵信号.模式识别模块是其重要组成部分, 实时智能区分安全扰动和危险入侵.本文提出一种用于光纤链路振动信号模式识别的复合特征提取方法.利用改进的双门限方法确定有效信号段的起止位置, 结合最大能量与最高信噪比挑选出采样周期内主要入侵扰动的特征段.综合利用特征段时域持续时间和小波包能量谱提取复合特征向量, 使用支持向量机进行模式识别.实验表明, 基于本文提出的规整化特征提取方法的模式识别准确率有了显著提高.     

基于多项式调频Fourier变换的信号分量提取方法

为了从含有噪声的混合信号中有效提取各个信号分量, 提出一种基于多项式调频Fourier变换的分量提取方法. 通过研究Fourier变换和分数阶Fourier变换的信号能量积累方式及变换基函数的时频表示, 提出利用时频平面上的多项式调频曲线族代替Fourier变换和分数阶Fourier 变换的调频直线族, 将变换的适用范围扩展到非线性调频信号. 采用粒子群智能优化算法搜索调频曲线族的最优多项式参数, 使混合信号中的某一分量在多项式调频Fourier域上能量谱集中. 最后对能量谱集中的分量进行窄带滤波, 并利用多项式调频逆Fourier变换重构信号分量. 仿真实验结果表明, 该方法不仅能够提取混合信号中的线性调频分量, 还能够实现非线性调频分量的能量谱集中、信号分离和时频特征提取.     

基于调频信号子带相关性分析的回波特征提取方法

利用两种典型调频信号(LFM及HFM)的波形特点,本文提出了一种基于目标回波子带相关性的特征分析方法,并提取了回波统计特征。仿真与海上试验数据分析结果表明,该方法提取的回波统计特征对于识别水下人造金属目标与非目标具有较好的可分性。     

基于无线传感器网络的地震信号特征提取方法研究

为解决野外古墓葬安防网络中高采样率会缩短无线传感器网络寿命的问题, 提出了使用功率谱二次处理对地震信号进行特征提取的方法. 并通过三类地面活动数据采集进行对比识别实验, 分析了低采样率条件下地震信号特征提取方法的性能. 结果表明, 使用功率谱二次分析的特征提取方法能够降低网络通信能耗, 延长网络寿命, 提高系统目标识别的准确性.该方法已应用于秦始皇兵马俑野外文物安防系统, 经实践检验, 收到了良好的效果. 关键词： 地震信号 特征提取 功率谱二次处理 无线传感器网络     

基于电子信号处理的四频激光陀螺信号读出研究

针对当前四频差动激光陀螺(DLG)读出系统结构复杂、易受温度影响等问题,对现有读出系统进行了改进,设计了使用电子信号处理代替光学信号处理来分离左、右旋陀螺信号的新型读出系统。通过对一个DLG同时使用新旧两套不同的读出系统进行测试,得到了相同的测试结果,表明电子信号处理和光学信号处理在分离左、右旋陀螺信号方面是等效的。在此基础上,首先使用模数转换器对DLG的光电转换信号进行采样,然后进行了数字信号处理,并研制出了全数字化读出系统。新型读出系统不使用四分之一波片和偏振片等光学元件,具有温度稳定性好、结构简单、全数字化等优点,有利于四频陀螺性能的提高和批量生产。     

基于经验模式分解的混沌干扰下谐波信号的提取方法

由混沌信号和谐波信号组合而成的复杂信号的分离方法一直受到关注.利用经验模式分解方法,依据任何信号由不同的固有简单振动模态组成的概念,将由混沌信号和谐波信号组合而成的复杂信号分离为不同的内在模态函数,并从中分解出谐波信号.通过利用Duffing方程产生的混沌信号进行的仿真实验,结果都表明该方法在一定参数范围内非常有效. 关键词： 经验模式分解 混沌 信号处理     

Optical time-domain differentiation based on intensive differential group delay

An optical time-domain differentiation scheme is proposed and demonstrated based on the intensive differential group delay in a high birefringence fibre waveguide. Results show that the differentiation waveforms agree well with the mathematically calculated derivatives. Both error and efficiency will increase when the birefringence fibre becomes longer, and the error rises up more quickly while the efficiency approaches to a maximum of ～0.25. By using a 1-m birefringence fibre a lower error of ～0.26% is obtained with an efficiency of 1% for the first-order differentiation of 10-ps Gaussian optical pulses, and the high-order optical differentiation up to 4th order is achieved with an error less than 3%. Due to its compact structure being easy to integrate and cascade into photonic circuits, our scheme has great potential for ultrafast signal processing.     

All-optical XOR logic gate operating on inputs with different modulation formats

An all-optical logic gate which can perform logic exclusive-OR (XOR) between an intensity modulated signal and a phase modulated one, is proposed and numerically investigated. The working principle of this logic gate is based on cross phase modulation in a highly nonlinear fiber.     

Cyclodextrin-functionalized magnetic graphene as solid-phase extraction absorbent coupled with flame atomic absorption spectrophotometry for determination of cadmium in water and food samples

A magnetic hydroxypropyl-β-cyclodextrin-functionalized reduced graphene oxide was synthesized and used as a magnetic solid-phase extraction adsorbent for the enrichment of cadmium from water and food samples before its determination by flame atomic absorption spectrometry. The effects of pH, amount of adsorbent, extraction time, desorption condition, and coexisting ions on the extraction efficiency were investigated and optimized. Under the optimal conditions, the limit of detection was 0.23?µg?L^?1 for water samples and 0.015?µg?g^?1 for food samples, respectively. The linear range was between 0.50 and 100.0?µg?L^?1 with a good correlation coefficient of 0.9986 in water samples, and between 0.050 and 5.0?µg?g^?1 with a correlation coefficient of 0.9975 in food samples. The method recoveries for the spiked samples were in the range of 87.5–102.4%, with the relative standard deviations ranging from 4.7% to 6.2% and enrichment factors ranging from 38 to 41, respectively. The adsorbent could be reused over 50 times without a significant change of extraction capability. The established method provides a promising environmentally friendly alternative for the determination of cadmium in complex matrix samples with no need of additional chelating reagents during the extraction process.     

A conservative level-set based method for compressible solid/fluid problems on fixed grids

A three-dimensional Eulerian method is presented for simulating dynamic systems comprising multiple compressible solid and fluid components where internal boundaries are tracked using level-set functions. Aside from the interface interaction calculation within mixed cells, each material is treated independently and the governing constitutive laws solved using a conservative finite volume discretisation based upon the solution of Riemann problems to determine the numerical fluxes. The required reconstruction of mixed cell volume fractions and cut cell geometries is presented in detail using the level-set fields. High-order accuracy is achieved by incorporating the weighted-essentially non-oscillatory (WENO) method and Runge–Kutta time integration. A model for elastoplastic solid dynamics is employed formulated using the tensor of elastic deformation gradients permitting the equations to be written in divergence form. The scheme is demonstrated using selected one-dimensional initial value problems for which exact solutions are derived, a two-dimensional void collapse, and a three-dimensional simulation of a confined explosion.     

Experimental investigation on the flame front resistance of gas channel growth with melt formation in iron ore sinter beds

The resistance of the flame front within the solid bed constitutes a fundamental and crucial area in porous bed combustion as the flame front propagation is highly related to the productivity and product quality. This paper focuses on the iron ore sintering, a thermal agglomeration process in steel mills. The results from a detailed experimental study of the pilot-scale pot tests under the conditions of a wide range of fuel rate are presented. The primary objective is to provide better understanding of the growth of gas channels relating to melt formation in the flame front and its resistance to flow. The sintering bed was divided into several zones based on the temperature profile and component distribution. Even though there is a continuous one-to-one replacement of humidified zone with porous sintered zone, a constant air flow rate during sintering could be obtained, indicating the ～100?mm high-temperature zone has a controlling effect on sintering bed permeability. The specific pressure drop value in high-temperature zone increases from ～3?kPa in upper bed to ～7?kPa in bottom bed, which varies with the bed temperature and structure properties. Both the green bed and sintered bed were scanned by X-ray computed tomography, the reconstruction and image analysis showed that the sintered bed has large gas channels and many more closed pores due to solid-melt-gas coalescence. More melt is generated when the heat is accumulated along the bed or input higher coke content, showing a propensity to suppress the gas channel growth and amplify the mismatch of gas transportation along the bed. Higher coke rate leads to a higher resistance in flame front, resulting in a slower flame front speed. These results are aimed to provide quantitative validation for improvements of a numerical sintering model in a future work.     

A simple method of contrast enhancement and extremum extraction for interference fringes

A method of interference fringe processing by finding the average and the extremum gray levels of image points within a fixed or varying window and then computing the modified image intensity with a simple algorithm is proposed. This method can yield a uniform brightness and high (1 or near to 1) contrast at the same time over the whole image area for an input pattern with low contrast and considerable luminance variation, when a relatively large window is used. It can also perform extremum extraction if a small window is employed. This technique is simple, effective and convenient for use, as has been verified by a series of examples of simulated and experimental fringe processing.     

基于分形几何边界提取的图像跟踪方法

为了满足现代武器装备对图像跟踪系统的实时性和跟踪可靠性的双重要求，针对武器装备所跟踪目标的特点，将基于分形几何的边界提取方法用于图像跟踪系统，提出了一种基于运动区域和边界图像的目标跟踪方案。文中的边界提取方法不仅能提取出目标的轮廓，还能较好地保留其内部细节，这样通过模板匹配得到的最佳匹配位置更加可靠。同时，基于运动区域的边界提取，尤其是对于远景的目标跟踪大大减少了计算量。此外，在利用边界图像进行匹配的基础上，采用了自适应模板更新策略，目的是尽可能地克服形变、光照等影响，使跟踪过程准确可靠。仿真实验表明，该方法运算量小，能满足武器装备实时性要求。     

基于信息融合的模拟电路故障的特征提取与融合方法

在模拟电路故障诊断中,故障特征的提取是一个非常重要的环节,其提取结果的好坏将直接影响最终的诊断正确率。对现有文献研究发现,每种特征提取方法单独使用时都有一定的局限性,为了能够更加充分的提取模拟电路故障特征,提出了小波包分析与主元分析并行应用的方法,并将两种方法提取的特征向量依据不同规则进行了三种类型的融合,方便对比实验。为获取最优小波特征,提出了特征偏离度,并以此为标准选择最优小波基。最后,通过设计一种改进的神经网络分类器模型,将融合后的三种特征向量送入其中进行仿真验证,得出最终诊断结果。结果表明,该方法能够有效克服单一特征提取方法提取不充分的缺点,提高故障诊断的正确率,并且融合因子 适中时诊断正确率最高。     

采用空间谱加权稀疏约束的稳健Capon波束形成方法

针对标准Capon波束形成器旁瓣级高以及期望信号方向存在误差时,性能会严重下降的问题,提出一种基于噪声及干扰空间谱加权的稀疏约束Capon波束形成方法。该方法利用波束响应本身具有的稀疏特性,以及在稀疏重构算法中l₁范数具有抑制较大值保护较小值的特性,使用噪声及干扰的空间谱对不同角度的波束响应进行加权,然后施以稀疏约束。仿真、水池试验与湖上试验中,与其它几种方法比较表明该方法可以降低波束形成器的旁瓣级,获得较深的零陷,提高了阵列输出信干噪比与抗导向向量误差的能力。     

基于位相抽取的三维信息加密算法研究

本文提出了以位相抽取为基础的三维信息加密算法.构造由纯振幅和纯位相物体组成的简单三维信息作为加密对象.先用标量衍射理论计算其复振幅并抽取相位分布,再结合双随机位相编码完成加密.解密算法为上述算法的逆过程.计算机模拟结果证实了该算法的有效性、鲁棒性和安全性,并揭示了位相抽取算法用于加密更大信息量三维信息的潜力. 关键词： 傅里叶光学 光学信息安全 三维信息加密 位相抽取     

基于图像处理的高精度透镜包边方法研究

提出了一种基于图像处理的透镜包边方法,根据待包边透镜的特点确定了光路中各参数,由面阵CMOS器件接收透镜反射的十字叉丝像,并通过图像处理的方法计算透镜的中心偏,以此为依据调整透镜位置,达到误差允许范围,然后加工金属外圆,完成包边。对测量系统、电机控制和软件实现作了详细说明,给出了实验结果。实验结果表明,该方法操作简单、测量精度高,可有效减小透镜中心偏。     

基于Hough投票空间的椭圆图像特征亚像素提取方法

椭圆检测是机器视觉和模式识别的基本任务。首先在椭圆中心初定位阶段,通过参数空间的少量投票,依据均值漂移技术获取粗略的椭圆中心,然后再由精提策略实现椭圆中心的精确像素级定位,并采用插值方法进一步获取椭圆中心亚像素坐标。仿真及真实图像的实验结果表明,所提出的方法同时具备了椭圆检测的鲁棒性和中心定位的精确性。