Complex, linear-phase filters for efficient image coding

With the exception of the Haar basis, real-valued orthogonal wavelet filter banks with compact support lack symmetry and therefore do not possess linear phase. This has led to the use of biorthogonal filters for coding of images and other multidimensional data. There are, however, complex solutions permitting the construction of compactly supported, orthogonal linear phase QMF filter banks. By explicitly seeking solutions in which the imaginary part of the filter coefficients is small enough to be approximated to zero, real symmetric filters can be obtained that achieve excellent compression performance     

一类具有线性相位的二重多小波的构造

给出由实单一紧支撑正交的小波构造二重正交多小波的方法。具体地,首先由实单一的紧支撑尺度函数构造出单一紧支撑正交对称的复尺度函数,再由构造出的复尺度函数去构造二重正交紧支撑多尺度函数,然后给出由二重尺度函数构造二重小波的显式公式。紧支撑正交的单一小波除Haar小波外不具有任何对称性,它用作滤波器不可能有线性相位,而由实单一紧支撑正交的尺度函数构造出的二重尺度函数却是对称的,对应的二重小波可以是对称或反对称的,从而使得这种小波在信号处理的过程中具有线性相位。最后给出相应的构造算例。     

Multiresolution Monogenic Signal Analysis Using the Riesz–Laplace Wavelet Transform

The monogenic signal is the natural 2D counterpart of the 1D analytic signal. We propose to transpose the concept to the wavelet domain by considering a complexified version of the Riesz transform which has the remarkable property of mapping a real-valued (primary) wavelet basis of L₂(R²) into a complex one. The Riesz operator is also steerable in the sense that it give access to the Hilbert transform of the signal along any orientation. Having set those foundations, we specify a primary polyharmonic spline wavelet basis of L₂(R²) that involves a single Mexican-hat-like mother wavelet (Laplacian of a B-spline). The important point is that our primary wavelets are quasi-isotropic: they behave like multiscale versions of the fractional Laplace operator from which they are derived, which ensures steerability. We propose to pair these real-valued basis functions with their complex Riesz counterparts to specify a multiresolution monogenic signal analysis. This yields a representation where each wavelet index is associated with a local orientation, an amplitude and a phase. We give a corresponding wavelet-domain method for estimating the underlying instantaneous frequency. We also provide a mechanism for improving the shift and rotation-invariance of the wavelet decomposition and show how to implement the transform efficiently using perfect-reconstruction filterbanks. We illustrate the specific feature-extraction capabilities of the representation and present novel examples of wavelet-domain processing; in particular, a robust, tensor-based analysis of directional image patterns, the demodulation of interferograms, and the reconstruction of digital holograms.     

Wavelets and recursive filter banks

It is shown that infinite impulse response (IIR) filters lead to more general wavelets of infinite support than finite impulse response (FIR) filters. A complete constructive method that yields all orthogonal two channel filter banks, where the filters have rational transfer functions, is given, and it is shown how these can be used to generate orthonormal wavelet bases. A family of orthonormal wavelets that have a maximum number of disappearing moments is shown to be generated by the halfband Butterworth filters. When there is an odd number of zeros at π it is shown that closed forms for the filters are available without need for factorization. A still larger class of orthonormal wavelet bases having the same moment properties and containing the Daubechies and Butterworth filters as the limiting cases is presented. It is shown that it is possible to have both linear phase and orthogonality in the infinite impulse response case, and a constructive method is given. It is also shown how compactly supported bases may be orthogonalized, and bases for the spline function spaces are constructed     

Despeckling of medical ultrasound images using Daubechies complex wavelet transform

The paper presents a novel despeckling method, based on Daubechies complex wavelet transform, for medical ultrasound images. Daubechies complex wavelet transform is used due to its approximate shift invariance property and extra information in imaginary plane of complex wavelet domain when compared to real wavelet domain. A wavelet shrinkage factor has been derived to estimate the noise-free wavelet coefficients. The proposed method firstly detects strong edges using imaginary component of complex scaling coefficients and then applies shrinkage on magnitude of complex wavelet coefficients in the wavelet domain at non-edge points. The proposed shrinkage depends on the statistical parameters of complex wavelet coefficients of noisy image which makes it adaptive in nature. Effectiveness of the proposed method is compared on the basis of signal to mean square error (SMSE) and signal to noise ratio (SNR). The experimental results demonstrate that the proposed method outperforms other conventional despeckling methods as well as wavelet based log transformed and non-log transformed methods on test images. Application of the proposed method on real diagnostic ultrasound images has shown a clear improvement over other methods.     

CONSTRUCTION METHODS AND ALGORITHM DESIGN OF WAVELET BASES

Based on the brief introduction of the principles of wavelet analysis, this paper gives a summary of several typical wavelet bases from the point of view of perfect reconstruction of signals and emphasizes that designing wavelet bases which are used to decompose the signal into a two-band form is equivalent to designing a two-band filter bank with perfect or nearly perfect property. The generating algorithm corresponding to Daubechies bases and some simulated results are also given in the paper.     

Theory and lattice structure of complex paraunitary filterbankswith filters of (Hermitian-)symmetry/antisymmetry properties

The theory of the real-coefficient linear-phase filterbank (LPFB) is extended to the complex case in two ways, leading to two generalized classes of M-channel filterbanks. One is the symmetric/antisymmetric filterbank (SAFB), where all filters are symmetric or antisymmetric. The other is the complex linear phase filterbank (CLPFB), where all filters are Hermitian symmetric or Hermitian antisymmetric and, hence, have the linear-phase property. Necessary conditions on the filter symmetry polarity and lengths for the existence of permissible solutions are investigated. Complete and minimal lattice structures are developed for the paraunitary SAFB and paraunitary CLPFB, where the channel number M is arbitrary (even or odd), and the subband filters could have different lengths. With the elementary unitary matrices in the structure of the paraunitary SAFB constrained to be real and orthogonal, the structure covers the most general real-coefficient paraunitary LPFBs. Compared with the existing results, the number of parameters is reduced significantly     

Boundary operation of 2-D nonseparable linear-phase paraunitary filter banks

This paper proposes a boundary operation technique of 2-D nonseparable linear-phase paraunitary filter banks (NS-LPPUFBs) for size limitation. The proposed technique is based on a lattice structure consisting of the 2-D separable block discrete cosine transform and nonseparable support-extension processes. The bases are allowed to be anisotropic with the fixed critically subsampling, overlapping, orthogonal, symmetric, real-valued, and compact-support properties. First, the blockwise implementation is developed so that the basis images can be locally controlled. The local control of basis images is shown to maintain orthogonality. This property leads a basis termination (BT) technique as a boundary operation. The technique overcomes the drawback of NS-LPPUFBs that the popular symmetric extension method is invalid. Through some experimental results of diagonal texture coding, the significance of the BT is verified.     

Theory of Dual-Tree Complex Wavelets

We study analyticity of the complex wavelets in Kingsbury's dual-tree wavelet transform. A notion of scaling transformation function that defines the relationship between the primal and dual scaling functions is introduced and studied in detail. The analyticity property is examined and dealt with via the transformation function. We separate analyticity from other properties of the wavelet such as orthogonality or biorthogonality. This separation allows a unified treatment of analyticity for general setting of the wavelet system, which can be dyadic or M-band; orthogonal or biorthogonal; scalar or multiple; bases or frames. We show that analyticity of the complex wavelets can be characterized by scaling filter relationship and wavelet filter relationship via the scaling transformation function. For general orthonormal wavelets and dyadic biorthogonal scalar wavelets, the transformation function is shown to be paraunitary and has a linear phase delay of $ omega /2$ in $[{0},2pi )$.      

Construction of sparse signal representations with adaptive multiscale orthogonal bases

We propose a new approach to construct adaptive multiscale orthonormal (AMO) bases of R^N that provide highly sparse signal representations. Our new multilayer AMO basis design produces a high proportion of small scale vectors. The basis vectors are built from small scale to large scales, layer by layer. For each layer, the basis vector maximizes a p-norm measure of sparsity. We compare the sparsity ratios SR (i.e. the percentage of negligibly small coefficients) obtained with AMO and Daubechies wavelet bases for seven families of piecewise smooth signals with randomly located discontinuities. The signals are composed of polynomial, sinusoidal and exponential pieces. In all cases, AMO bases produce a SR increase ranging from 6% to 37%. AMO bases have three main advantages over wavelets. First, they are found automatically by solving a sequence of optimization problems, which eliminates the problem of selecting a wavelet for a given signal. Second, they can provide a significantly sparser representation. Finally, they have the ability to produce zero coefficients for a larger family of piecewise smooth signals. The drawbacks of AMO bases are computational: the basis computation is more expensive, the basis vectors require storage space and no fast transform is known.     

A new class of orthonormal symmetric wavelet bases using a complexallpass filter

This paper considers the design of the whole sample symmetric (WSS) paraunitary filterbanks composed of a single complex allpass filter and gives a new class of real-valued orthonormal symmetric wavelet bases. First, the conditions that the complex allpass filter has to satisfy are derived from the symmetry and orthonormality conditions of wavelets, and its transfer function is given to satisfy these conditions. Second, the paraunitary filter banks are designed by using the derived transfer function from the viewpoints of the regularity and frequency selectivity. A new method for designing the proposed paraunitary filterbanks with a given degrees of flatness is presented. The proposed method is based on the formulation of a generalized eigenvalue problem by using the Remez exchange algorithm. Therefore, the filter coefficients can be easily obtained by solving the eigenvalue problem, and the optimal solution is attained through a few iterations. Furthermore, both the maximally flat and minimax solutions are also included in the proposed method as two specific cases. The maximally flat filters have a closed-form solution without any iteration. Finally, some design examples are presented to demonstrate the effectiveness of the proposed method     

利用镜像传输实现OFDM载波间干扰抵消的方法

OFDM系统中由于存在载波频偏以及多普勒频移,使得系统易受ICI干扰,文中提出了两路分集镜像对称发送方式以降低OFDM系统的ICI干扰,该方法基于两路分集相位旋转共轭抵消方法,采用两路镜像对称的信号代替共轭对称信号,通过接收机的还原处理自动引入相位旋转因子,与相位旋转共轭对称方式相比,该方法避免了频偏估计以及回传处理,降低了系统复杂性,提升了系统传输效率。     

具有任意正则阶的M-带对称正交小波设计

利用计算代数中Grobner基与合冲模的概念与算法,论文提出了一种多相位矩阵的正交化方法,在此基础上得到了同时具有对称性和任意正则阶的M-带正交小波的高效设计方法.克服了现有算法构造过程复杂以及不能保持线性相位的缺陷,另外,当所求得的尺度滤波器系数为参数形式时,本文算法求得的小波滤波器也为含参数的,因而可以根据实际问题灵活选取适当参数从而得到所需要的M-带小波系统.     

Matching pursuit and atomic signal models based on recursive filterbanks

The matching pursuit algorithm can be used to derive signal decompositions in terms of the elements of a dictionary of time-frequency atoms. Using a structured overcomplete dictionary yields a signal model that is both parametric and signal adaptive. In this paper, we apply matching pursuit to the derivation of signal expansions based on damped sinusoids. It is shown that expansions in terms of complex damped sinusoids can be efficiently derived using simple recursive filter banks. We discuss a subspace extension of the pursuit algorithm that provides a framework for deriving real-valued expansions of real signals based on such complex atoms. Furthermore, we consider symmetric and asymmetric two-sided atoms constructed from underlying one-sided damped sinusoids. The primary concern is the application of this approach to the modeling of signals with transient behavior such as music; it is shown that time-frequency atoms based on damped sinusoids are more suitable for representing transients than symmetric Gabor atoms. The resulting atomic models are useful for signal coding and analysis modification synthesis     

Shaped-pattern synthesis using pure real distributions

Shaped patterns can be produced by properly excited equispaced linear arrays. An earlier synthesis procedure, which accomplishes this with control ripple in the shaped region and controlled sidelobe levels elsewhere, results in array distributions that are generally complex. It is shown here that if the shaped pattern is symmetric and has 2M filled nulls, there are 2^M complex symmetric distributions, and 2^M pure real asymmetric distributions, and 2^2M-2^M+1 complex asymmetric distributions that will produce the desired pattern. By adding 2M elements to the array, one can find a symmetric pure real distribution that will achieve the same result. A representative example illustrates the procedure. The results have application to standing-wave-fed planar arrays with quadrantal symmetry via use of the collapsed distribution principle     

Three-band biorthogonal interpolating complex wavelets with stopband suppression via lifting scheme

Wavelet research has primarily focused on real-valued wavelet bases. However, the complex filterbanks provide much convenience for complex signal processing. For example, in radar and sonar signal processing, the complex signals from the I/Q receiver can be efficiently processed with complex filterbanks rather than real filterbanks. Specifically, the positive and negative Doppler frequencies imply different physical content in the moving target detector (MTD) and moving target identification (MTI); therefore, it is significant to design complex multiband filterbanks that can partition positive and negative frequencies into different subbands. We design two novel families of three-band biorthogonal interpolating complex filterbanks and wavelets by using the three-band lifting scheme. Unlike the traditional three-band filterbanks, the novel complex filterbank is composed of three channels, including the lowpass channel, the positive highpass channel whose passband distributes in the positive frequency region, and the negative highpass channel in the negative frequency region. Such a filterbank/wavelet naturally provides the ability to extract positive frequency components and negative frequency components from complex signals. Moreover, a novel set of design constraints are introduced to manipulate the stopband characteristic of highpass filters and are referred to as stopband suppression, which strengthens the traditional constraints of vanishing moments. Finally, a numerical method is given to further lower stopband sidelobes.     

Pascal's triangle: An origin of Daubechies polynomials and an analytic expression for associated filter coefficients

After showing that Daubechies polynomial coefficients can be simply obtained from Pascal's triangle by some elementary additions, we propose a derivation of the spectral factorization by using the elementary symmetric functions. This derivation leads us to present an analytic expression, able to compute Daubechies wavelet filter coefficients from the roots of the associated Daubechies polynomial. Thus, these coefficients are directly obtained and without recurrence. At last, we measure the quality of the coefficient sets generated by this expression and we compare it with two well-known methods.     

基于多小波分解的多光谱图像矢量融合

在实数域中,对称、正交的紧支集非平凡单小波基不存在,而多小波把紧支性、对称性、正交性完美地结合在一起,使小波理论从标量扩展到矢量范畴。考虑到图像多小波变换系数具有矢量特性,该文将基于像素点和基于区域的标量融合策略推广到矢量情形,提出一种新的、在多小波域中基于矢量融合的图像融合算法,充分利用多小波变换域系数矢量内部各个分量的相关性来提高融合质量。两波段真实多光谱图像融合实验结果表明,与单小波标量融合方法相比,多小波矢量融合算法获得的图像具有较优的视觉效果和客观评价指标,从而证明了用于图像融合时,多小波较之单小波更适合于人类视觉系统,具有广泛的应用前景。     

矩量法与C_（3V）群基函数求解二维电磁散射的应用

根据散射体的几何对称特性,将群的对称理论和基函数的选取与矩量法相结合,可以有效减少计算存储量和计算时间。对于横截面为正三角的散射柱体,取正三角形各顶点和各边中点为对称点,构造6个自然对称基函数。将正规表示基函数约化为不可约表示,进而确定出上述6个自然对称基的线性组合的独立完备基。结果表明,以该基函数级数展开计算具有收敛速度快,计算精度高等特点。     

Compressed sensing of complex-valued data

Compressed sensing (CS) is a recently proposed technique that allows the reconstruction of a signal sampled in violation of the traditional Nyquist criterion. It has immediate applications in reduction of acquisition time for measurements, simplification of hardware, reduction of memory space required for data storage, etc. CS has been applied usually by considering real-valued data. However, complex-valued data are very common in practice, such as terahertz (THz) imaging, synthetic aperture radar and sonar, holography, etc. In such cases CS is applied by decoupling real and imaginary parts or using amplitude constraints. Recently, it was shown in the literature that the quality of reconstruction for THz imaging can be improved by applying smoothness constraint on phase as well as amplitude. In this paper, we propose a general l_p minimization recovery algorithm for CS, which can deal with complex data and smooth the amplitude and phase of the data at the same time as well has the additional feature of using a separate sparsity promoting basis such as wavelets. Thus, objects can be better detected from limited noisy measurements, which are useful for surveillance systems.