基于可微收缩函数与自适应阈值的小波去噪

为进一步提升高斯噪声的去噪效果,提出了基于可微收缩函数与自适应阈值的图像去噪方法.根据高斯噪声的小波系数具有幅值小、服从高斯分布的特征,提出一种自适应于信噪强度的阈值,以准确地区分噪声系数与图像系数.根据自然图像的小波系数具有平滑连贯的特征,提出了一种可微的收缩函数,与自适应阈值结合对含噪的小波系数进行量化处理,以有效...     

基于小波变换的图像混合噪声自适应滤除算法

为同时滤除图像中的椒盐噪声和高斯噪声,提出了一种基于小波变换的混合噪声自适应滤除算法,该算法首先采用中值滤波去除椒盐噪声,然后借助边缘检测算子区将图像为分边缘与非边缘区域,进一步对非边缘区域引入改进的均值滤波器,有效削弱高斯噪声的同时保护图像边缘细节,既初步削弱高斯噪声又保护了边缘,最后采用改进的小波阈值滤波算法,对不同的小波系数采用不同的阈值函数,通过线性回归得到各最优阈值关系式。实验结果表明,该混合噪声自适应滤除算法能有效滤除椒盐噪声和高斯噪声,在图像主观质量和客观质量上均取得了较好的效果,能提高去噪图像峰值信噪比0.5~2.0 dB。     

镜头边界检测的动态窗口技术

提出了一种新的镜头边界检测算法.该算法借鉴信息论的观点,以互信息量和直方图的结合来衡量帧间相似度.通过一自动选取的阈值选出所有可能的镜头切换点,采用动态窗口技术将切变和渐变统一视为切变检测.通过合理的聚类消除噪音所带来的误检现象,提高了镜头的检出率.仿真实验结果表明该算法具有良好的性能.     

基于EM和GMM相结合的自适应灰度图像分割算法

提出一种阈值自适应、EM方法估计GMM参量的图像分割算法,能够根据图像的内容结合区域和边界两方面的信息自适应地选择阈值,精确地进行图像边界分割.算法首先提取图像的边界,然后根据边界的直方图计算图像的可分割性,由可分割性确定EM方法的阈值进行GMM分割,最后合并图像的近似区域.实验数据表明,相比其它图像分割算法,以及固定阈值的传统EM算法,本算法的分割结果更为准确.     

一种改进的基于混合高斯模型的运动目标检测方法

混合高斯模型方法被广泛应用于运动目标检测,但是现有的混合高斯模型方法在应对噪声和光照突变时效果不佳,其运动目标检测的效果会受到严重影响。为了解决上述问题,提出了一种结合边缘混合高斯模型方法以及改进的基于邻域差方法的综合方法。该方法充分利用了边缘图像对于噪声和光照突变不敏感,以及邻域差方法可以去除部分噪声的特点,对图像序列进行综合处理。实验结果证明,该方法可以提升运动目标的检测率,降低误警率,可以更有效地应对噪声和光照突变的干扰,从而具有更准确的目标检测效果。     

基于高阶累计量的大气光通信自适应信号处理

针对强湍流信道下信号衰落的特点,分析了对数正态分布模型与K分布模型的适用范围.基于K分布模型建立大气光通信接收信号模型,并给出了自适应最优门限检测方法.采用四阶和六阶累计量对强湍流信道参量进行估计,采用二阶累计量对其它高斯噪音进行估计,得到K分布参量及高斯噪音统计量的预测值,实现自适应门限更新.基于Monte Calro算法进行仿真,给出了门限更新算法对通信系统误码率的影响,同时分析了信号采样率对估计参量偏差的影响.计算表明,在强湍流情况下,大气光通信系统的误码率性能得到极大的改善,优于基于MLSD检测的接收机.     

基于混沌粒子群优化的图像Contourlet阈值去噪

提出了基于混沌粒子群优化的图像Contourlet阈值去噪方法.该方法在Contourlet变换域内利用混沌粒子群算法来确定最优阈值,再通过软阈值函数去噪,且不需要噪音方差等先验信息.实验结果表明:该方法与小波Bayeshrink阈值、基于粒子群的小波阈值、Contourlet自适应阈值等去噪方法相比,能有效地去除高斯白噪音和椒盐噪音的混合噪音,提高峰值信噪比,并较好地保留图像的细节和纹理,从而明显地改善了图像的视觉效果.     

基于混沌粒子群优化的图像Contourlet阈值去噪

提出了基于混沌粒子群优化的图像Contourlet阈值去噪方法.该方法在Contourlet变换域内利用混沌粒子群算法来确定最优阈值,再通过软阈值函数去噪,且不需要噪音方差等先验信息.实验结果表明：该方法与小波Bayeshrink阈值、基于粒子群的小波阈值、Contourlet自适应阈值等去噪方法相比,能有效地去除高斯白噪音和椒盐噪音的混合噪音,提高峰值信噪比,并较好地保留图像的细节和纹理,从而明显地改善了图像的视觉效果.     

基于多特征检测与支持向量回归的图像文本提取算法

为解决复杂背景中难以有效提取场景文本的问题,提出了一种基于多特征检测与支持向量回归的图像文本提取方案。为有效区分文本与非文本边缘,基于图像边缘,提取场景中三个文本特征。将得到的三个文本特征进行多尺度融合,利用文本融合特征检测候选文本边界,有助于检测不同大小的文本,提高对不同类型的图像退化的鲁棒性。对于每个检测到的候选文本边界,根据邻域窗口中的像素来估计每个像素的局部阈值,利用局部阈值自适应分割提取候选字符。引入支持向量回归模型对文本像素与图像背景精确分离,消除非文本边界,提取真实字符和单词。实验表明:与当前文章提取技术相比,所提方法具有更好的鲁棒性,能适用各种变化的复杂场景文本提取,具有更优的PrecisionRecall曲线与F测量值。     

基于自适应侧抑制网络的红外弱小目标检测

 为了提高远距离红外弱小目标的检测效率,提出了一种基于自适应侧抑制网络的复杂背景下的红外弱小目标检测方法。该方法建立了改进的侧抑制网络数学模型,利用各向异性滤波来自适应地确定侧抑制网络的抑制系数,不需要人为干预,实现了侧抑制网络与各向异性高斯滤波的有机结合。同时,在各向异性高斯滤波器两轴确定方面进行了改进,两轴分别采用对比度尺度模型和强度尺度传播模型来独立确定。通过与传统弱小目标检测方法的对比实验,验证了方法的有效性。     

Optical theorems and steinmann relations

Formulas that express in terms of physical scattering functions the discontinuity of any 3-to-3 scattering function across any basic normal threshold cut are derived from field theory. These basic cuts are the cuts in channel energies that start at lowest normal thresholds and extend to plus infinity. The discontinuity across such a cut generally depends on whether it is evaluated above or below each of the remaining basic cuts. Formulas are obtained for all cases. Generalized Steinmann relations are found to hold: the 2282 boundary values from which the discontinuities across basic cuts are formed have a unique extension to a set of 2¹⁶ = 65,536 functions, one for each combination of sides of the 16 basic cuts, such that for any pair of overlapping channels the corresponding double discontinuity vanishes. The ordinary Steinmann relations require this property to hold only for the double discontinuities formed from the original 2282 boundary values. The results are derived from the field-theoretic formalism of Bros, Epstein, and Glaser, which is slightly developed and cast into a form suited for calculations of the kind needed here.     

建筑构件内瞬态热湿分析中的多层求解方法

提出了一种求解一维多层建筑构件内瞬态耦合传热传质问题的有限差分逼近方法.求解过程中认为构件的内外表面处的热湿传递为瞬态过程,并考虑了多层墙体每一层物性参数的非连续性,从而避免了通常处理中边界条件设定为常数给计算带来的误差.利用该方法对典型的三层墙体层与层界面处相对湿度在24小时内的变化进行了计算,计算结果与文[6]的传递函数求解方法符合很好,可推广应用到对二维和三维耦合传热传湿问题的求解.     

Sources of spurious force oscillations from an immersed boundary method for moving-body problems

When a discrete-forcing immersed boundary method is applied to moving-body problems, it produces spurious force oscillations on a solid body. In the present study, we identify two sources of these force oscillations. One source is from the spatial discontinuity in the pressure across the immersed boundary when a grid point located inside a solid body becomes that of fluid with a body motion. The addition of mass source/sink together with momentum forcing proposed by Kim et al. [J. Kim, D. Kim, H. Choi, An immersed-boundary finite volume method for simulations of flow in complex geometries, Journal of Computational Physics 171 (2001) 132–150] reduces the spurious force oscillations by alleviating this pressure discontinuity. The other source is from the temporal discontinuity in the velocity at the grid points where fluid becomes solid with a body motion. The magnitude of velocity discontinuity decreases with decreasing the grid spacing near the immersed boundary. Four moving-body problems are simulated by varying the grid spacing at a fixed computational time step and at a constant CFL number, respectively. It is found that the spurious force oscillations decrease with decreasing the grid spacing and increasing the computational time step size, but they depend more on the grid spacing than on the computational time step size.     

基于不连续因子的沸水堆芯三维瞬态数值模拟

基于不连续因子校正的粗网格有限差分法是实现堆芯瞬态三维数值模拟的高效方法之一, 粗网节块的界面不连续因子与边界反照率的计算方法决定了实时数值模拟过程中的精度。在计算不连续因子的过程中, 省去了细网节块计算与粗网均匀化过程, 直接在粗网格划分情况下, 基于节块展开法和非线性迭代策略, 推导了粗网格界面不连续因子比率与边界反照率的计算公式, 并编制了相应的计算程序。沸水堆典型算例的三维瞬态模拟证实该方法可在空间域和时间域两方面, 使静态、瞬态精度均达到与先进节块法相等同的程度, 并且计算效率优于先进节块法, 为核电站全范围模拟机三维堆芯的实时仿真模型开发提供了一种切实可行的选择。     

Time-domain impedance formulation for transmission line matrix modelling of outdoor sound propagation

Outdoor sound propagation modelling has attracted considerable attention in the past and has lead to many analytical and numerical models. More recently, the increase of computational power has led to consider time-domain methods that enable to consider transient phenomena. Among these models, the transmission line matrix (TLM) method has been proposed, but the sound absorption at boundaries appears to be a somewhat underdeveloped aspect of this approach. In this paper, a specific implementation of impedance boundary condition is proposed. The method is based on the approximation of the impedance as a sum of linear systems, which allows the formulation of an equivalent impedance model in the time-domain. The proposed implementation is applied for two common impedance models of porous material. Numerical simulations have been carried out in the case of sound propagation over a flat ground with and without an impedance discontinuity, and for several values of specific airflow resistivity. TLM numerical predictions expressed in terms of excess attenuation relative to free field show a good agreement with analytical solutions.     

相界上运动位错的弹性场

本文推广Eshelby等人关于无限各向异性介质中静止直线位错的处理方法,提出了计算两相介质中相界面上匀速运动位错弹性场的方案。本文所提出的简便方法有一般的适用性,可直接用于研究在相界面上高速运动的相变位错的行为,从而有助于理解位错运动在无扩散相变过程中的作用。最后对本文的方法及观察到的场中的间断面作了简短讨论。 关键词：     

Scattering by cavities of arbitrary shape in an infinite plate and associated vibration problems

This paper presents a solution for the displacement of a uniform elastic thin plate with an arbitrary cavity, modelled using the biharmonic plate equation. The problem is formulated as a system of boundary integral equations by factorizing the biharmonic equation, with the unknown boundary values expanded in terms of a Fourier series. At the edge of the cavity we consider free-edge, simply-supported and clamped boundary conditions. Methods to suppress ill-conditioning which occurs at certain frequencies are discussed, and the combined boundary integral equation method is implemented to control this problem. A connection is made between the problem of an infinite plate with an arbitrary cavity and the vibration problem of an arbitrarily shaped finite plate, using the jump discontinuity present in single-layer distributions at the boundary. The first few frequencies and modes of displacement are computed for circular and elliptic cavities, which provide a check on our numerics, and results for the displacement of an infinite plate are given for four specific cavity geometries and various boundary conditions.     

基于区域合并的动态阈值分割算法

在传统分块分割再合并阈值法的基础上,引入图像区域合并的思想,以邻近区域阈值相似性作为判断区域合并的依据,借助人机交互式系统,运用待识别目标灰度与完成区域合并的阈值作加权,消除区域灰度不连续,实现目标与背景的有效分离,减少算法的运算量,提高目标识别精度.实验证明,该算法实现简单且行之有效.     

On the boundary conditions in crystal optics with spatial dispersion

The non-homogeneous boundary conditions in the theory of spatial dispersion are considered. It is shown that non-homogeneity in the boundary condition for polarization causes a discontinuity in the tangential components of the magnetic field at the boundary. Separate boundary conditions do not contradict the Energy Conservation Law only if they are connected in a certain way. The values of parameters introduced in the theory are obtained with the help of numerical methods. The theoretical spectra of reflectivity obtained with such parameters coincide with the experimental ones.     

Instability of a charged plane interface between two liquids with respect to the tangential discontinuity of a time-dependent velocity field

The differential equation that describes the evolution of perturbations of a charged plane boundary between immiscible liquids when the upper liquid moves relative to the lower one with a time-dependent velocity parallel to the boundary is the Hill equation. In this system, the interface can exhibit instabilities of three types at various values of physical parameters: the Kelvin-Helmholtz, Tonks-Frenkel, and parametric instability. When physical parameters have certain values, the interface that is unstable with respect to surface charge and the tangential discontinuity of the velocity field across the interface can be parametrically stabilized.