A multidimensional wave digital filter bank for video-based motion analysis

Multidimensional wave digital filters (MDWDF) exhibit the same desirable properties as 1D WDFs, most notably including passivity and therefore guaranteed stability as well as high robustness. A possible application for such MDWDFs may be found in motion analysis of image sequences by means of filters with fan-shaped transfer functions, where content with specific movement information can be extracted. For that matter, a parallel filter bank is needed to differentiate object motion into separate classes. In this paper, a new specialized MDWDF fan filter structure is introduced, possessing both reduced computational complexity and memory requirements compared to existing approaches. Additionally, part of the processing can be shared among all bands, further increasing efficiency.     

Exploiting the parallelism of multidimensional wave digital algorithms on multicore computers

Multidimensional wave digital algorithms for numerical integration of partial differential equations exhibit not only important robustness properties, but also a massive amount of parallelism. As the technology limit of heat dissipation stalls a further increase of clock rates, modern CPUs incorporate multiple cores for parallel computation. In this paper, a safe and efficient multithreading concept is presented to exploit the multicore architecture for multidimensional wave digital algorithms. Context switching and synchronization overhead is investigated as well as effects of unfair operating system thread scheduling due to unequal cache sharing of cores. Simulation results for the nonlinear Euler equations confirm the efficiency of the proposed setup on a 1-core, 4-core and a 2 × 4-core system.     

Serial block pipelined wave digital filter

Based on serialising a previously established block pipelined architecture for PCAS wave digital filters, a lower transistor count, lower power implementation is derived     

Vector one-way wave absorbing boundary conditions for FEMapplications

In the paper a derivation is presented which leads to a new and general class of vector absorbing boundary conditions (ABCs) for use with the finite element method (FEM). The derivation is based on a vector one-way wave equation and a polynomial approximation of the vector radical. It is shown that wide-angle absorbing boundary conditions, as proposed in Halpern and Trefethen (1988) for optimal absorption of out-going waves, can be obtained in vector form. Vector plane waves are used to evaluate the accuracy and the reflection performance of these boundary conditions in a wide range of incidence angles. The implementation of the vector ABCs in a FEM formulation is also provided to show how up to the fifth-order absorbing accuracy can be achieved with derivatives only up to the second-order. A possible formulation is described which not only yields a third-order accuracy with first-order derivatives, but also retains the symmetry of the FEM matrix     

Phase equalization of multidimensional wave digital filters derived by rotation

In multidimensional filtering, the phase response is very important. It should be as linear as possible in a certain sense. This paper presents a numerical approach to equalize the phase of multidimensional wave digital filters whose reference filters are transformed from one-dimensional reference filters. This method is illustrated by an example of a two-dimensional wave digital filter whose transfer function has nearly circular symmetrical behavior. The parameters of the all-pass circuits for phase equalization are calculated.     

基于MATLAB的FIR滤波器的设计

基于MATLAB信号处理工具箱,设计一种滤波器,针对信号的分离和滤除,要求有良好的幅频特性和线性相位特性,且可通过参数来改变其主瓣宽度和旁瓣衰减。提出基于窗函数法设计滤波器,加入信号源,利用FDATool设计和分析模块电路,经过Simulink仿真,对其进行时域、频域分析和滤波。实验结果表明,本设计方法能够有效的滤除无...     

COMPUTER-AIDED ANALYSIS OF WAVE DIGITAL FILTER TRANSFER CHARACTERISTIC

An approach for finding the wave digital filter transfer characteristic is developed with ageneral computer program.The program can be used to analyse the quantification effect of differentmultiplier coefficient wordlengths for the wave digital filter transfer characteristic,thus it is possible to make areasonable choice of coefficient word-length in order to take account of economization as well as technicalspecifications when the wave digital filter is in hardware implementation.     

基于免疫算法的IIR数字滤波器优化与实现

由于IIR数字滤波器设计实质上是一个非线性高维复杂函数优化问题,文中提出基于具有全局搜索能力强,收敛速度快特点的免疫算法实现IIR数字滤波器优化设计的新方法,给出了IIR滤波器优化设计的数学模型,描述了应用免疫算法优化设计IIR数字滤波器的具体实现步骤。通过低通和高通IIR数字滤波器设计的仿真结果表明方法的有效性和高效性。     

Technique for flexible multidimensional filter design

This new design technique utilises one-dimensional half-band filter units arranged in a tree structure to split multidimensional frequency space into sections. These sections may be switched in or out to form an approximation to any desired frequency response. The technique has proved useful in television bandwidth reduction experiments using subsampling.     

Absorbing boundary conditions on arbitrary boundaries for thescalar and vector wave equations

The solution of open region scattering problems involving inhomogeneous arbitrarily shaped objects may be performed through the use of partial differential equation techniques, which require enclosing the scatterer by an outer boundary on which an absorbing boundary condition (ABC) is applied. In order to minimize the size of the domain to be meshed and, consequently, the number of unknowns, if may be advisable to implement ABC's devised for outer boundaries of arbitrary shapes. Such ABC's are obtained for the 3D scalar and vector wave equations; they incorporate most of existing boundary conditions. When used in conjunction with a finite element technique, the numerical results derived by using a simplified form of these ABC's compare favourably to those obtained by using a rigorous hybrid finite element-integral equation formulation. These boundary conditions have been obtained in the frequency-domain framework; they may, however, be used in time-domain calculations     

基于遗传算法的数字图像相关变形初值估计

为了克服传统的基于牛顿-拉夫森迭代的数字图像相关法受迭代初值影响较大等问题, 提出了一种结合遗传算法的数字图像相关法。以待测数据点为中心, 选取邻域内的若干估值点, 通过基于遗传算法的数字图像相关法匹配出变形前后估值点对坐标; 随机选取不共线的3组或以上估值点对代入仿射变换模型, 依据仿射变换结果估计变形初值, 并作为牛顿-拉夫森迭代初值; 最后结合牛顿-拉夫森迭代法计算亚像素位移值。结果表明, 该方法的匹配时间相对传统方法平均降低37.54%, 相较于传统的数字图像相关法在搜索性能、匹配精度等方面更加可靠。该研究为数字图像相关法中迭代初值优化效果对匹配速度和精度的影响提供了参考。     

Integrated high bandwidth wave elliptic lowpass switched-current filter in digital CMOS technology

The design and fabrication of a high bandwidth third-order analogue elliptic lowpass wave filter is described. The filter was realised using switched-current techniques and implemented on a standard digital 0.6 /spl mu/m CMOS process. The filter exhibits an elliptic response and achieves a bandwidth of 250 kHz, which is significantly higher than a previously reported fabricated switched-current wave filter.     

Numerical integration of partial differential equations using principles of multidimensional wave digital filters

Physical systems described by partial differential equations (PDEs) are usually passive (due to conservation of energy) and furthermore massively parallel and only locally interconnected (due to the principle of action at proximity, as opposed to action at a distance). An approach is developed for numerically integrating such PDEs by means of algorithms that offer massive parallelism and require only local interconnections. These algorithms are based on the principles of multidimensional wave digital filtering and amount to directly simulating the actual physical system by means of a discrete passive dynamical system. They inherit all the good properties known to hold for wave digital filters, in particular the full range of robustness properties typical for these filters.In this paper, only the linear case is considered, with particular emphasis on systems of PDEs of hyperbolic type. The main features are explained by means of an example.     

Perfect reconstruction with critically sampled filter banks andlinear boundary conditions

This work is concerned with the (image) boundary conditions involved in processing a finite discrete-time signal with a critically sampled perfect reconstruction filter bank. It is desirable that the boundary conditions reduce edge effects and define a transformation into a space having the same dimensionality as the original signal. The complication that arises is in the computation of the inverse transform. Although it is straightforward to reconstruct the signal values that were not influenced by the boundary conditions, recovering those values on the boundaries is nontrivial. The solution of this problem is discussed for general linear boundary conditions. No symmetry assumptions are made on the boundary conditions or on the impulse responses of the analysis filters. A low-rank linear transform is derived that expresses the boundary values in terms of the transform coefficients, which in turn provides a method for inverting the subband decomposition. The application of the results in the case of two-channel orthonormal wavelet filters is discussed, and the effects of the filter support on the conditioning of the inverse problem are investigated     

彩色数字全息常见波面重建算法的实现与比较

为了研究彩色数字全息检测应用中波面重建方法对重建场质量与计算速度的影响,及不同波长记录的重建场的准确融合,采用理论分析及实验验证的方法,给出了不同色光物光重建场准确重叠的方法及彩色数字全息波面实时再现的详细过程,并分析比较了几种波面重建算法的特点及计算速率。结果表明,使用球面波为重建光的可变放大率波面重建算法能适应各种大小的物体,且可获得较高分辨率的像,并占用较少的计算时间,能较好适用于彩色数字全息检测应用。     

A general scrambling rule for multidimensional FFT algorithms

This work determines the scrambling rule of the multidimensional Cooley-Tukey FFT, and of the multidimensional prime factor FFT, in complete generality, i.e., for signals defined on lattices of general type. The characteristics of the scrambling rule bear interesting similarities with the 1-D case: the scrambling can be performed on the input data and it can be eliminated from the operations requiring pairs of FFT and inverse FFT (e.g. convolutions and correlations). The results of this work allow one to derive the most efficient way of performing multidimensional scrambling. The consequent memory access savings are relevant, especially with arrays of sizable dimensions     

Synthesis and analysis of algorithms for digital signal recognition in conditions of deforming distortions and additive noise

The problem of digital signal recognition has been considered in conditions of deforming distortions of the waveform of these signals and additive Gaussian noise. A mathematical model for introducing deformations of the known or random waveform signals is proposed for synthesizing recognition algorithms. The model is based on introducing the nonlinear deformation operator as an operator of permutations with repetitions of elements of the initial discrete signal with addition of additive noise component caused by quantization errors of continuous deformation function. Two recognition algorithms were synthesized and investigated. The first is an optimal one based on the exact calculation of likelihood functions, and the second is a quasi-optimal algorithm based on using the Gaussian approximation of likelihood functions. These algorithms were simulated for different variants of the specified values of deforming distortions in the form of determinate functions and in the form of random function realizations. The experimental error probability was compared with its theoretical estimate at different values of signal-to-noise ratio.     

Absorbing boundary conditions applied to model wave propagation inmicrowave integrated-circuits

A general expression of an absorbing boundary condition is presented in this paper to model wave propagation in passive microwave integrated-circuits by the finite-difference time-domain method. Unlike previously developed absorbing boundary conditions which can only absorb propagating waves, this boundary condition can also absorb evanescent waves effectively. The microstrip line is used as an example to demonstrate how to impose this absorbing boundary condition on different outer boundaries of a computation-domain. It is also demonstrated that the numerical stability of this absorbing boundary condition, when it is applied in its high order form, can be maintained by properly selecting its parameters     

Design automation with the TSPC circuit technique: ahigh-performance wave digital filter

In this paper, we demonstrate how the true single-phase clocking (TSPC) circuit technique is utilized fur a high-speed recursive filter application, with a high degree of design automation. This features a quick netlist generation of integral high-speed arithmetic modules, utilization of carry-save architectures, and synthesis and optimization with a TSPC cell library. Implementation results in a 0.8 μm standard CMOS process indicate substantial performance improvements over traditional designs, at the same time keeping design time very short. Fabricated samples of the third-order lattice wave digital filter were measured at 265 Msamples/s, which is more than double the sample rate reported in previous works on the same filter and same or comparable technology