Asymptotic analysis of mutual coupling in concave circularcylindrical arrays. II. Near-zone

For pt.I, see ibid., vol.41, no.2, p.121-36 (1993). This work evaluates the near-zone coupling coefficients in large concave arrays. A canonical model of a circular cylindrical concave array of open-ended rectangular waveguides is analyzed. It is shown that in the paraxial region (which also includes the area close to the excited element) the coupling coefficients are identical, to lowest asymptotic order, to those in an equivalent infinite planar array. In the transition region, which lies between the paraxial and the far zone, or the ray region, two representations were developed: one, a transition function, expressed in terms of a canonical integral which must be numerically evaluated, and the other a uniform representation, which consists of a superposition of a planar array contribution and a few periodic structure rays. The uniform representation is valid in the near zone (which includes both the paraxial and the transition region) as well as in the far zone. This form is simple and may be immediately generalized to concave arrays with slowly varying curvature and periodicity     

基于高斯比例混合模型的图像非下采样Contourlet域去噪

为改善图像的去噪效果,该文提出了一种基于高斯比例混合模型的图像非下采样Contourlet域去噪算法。该算法首先建立非下采样Contourlet系数邻域的高斯比例混合模型,然后在模型基础上应用贝叶斯最小二乘法对系数进行估计,最后反变换得到恢复图像。算法结合了非下采样Contourlet变换对图像边缘的高效表示能力、非下采样变换的移不变性质以及GSM模型对非下采样Contourlet系数邻域相关性的概括能力。实验结果表明,该算法在视觉效果和峰值信噪比的改善上都取得了非常好的效果。     

基于多分辨率格网的三维物体识别方法

本文首先提出了一种改进的三维物体表达方法,它将一个三维物体表面网格与其它表面网格的几何关系表示为一个二维矩阵,称为距离角度图.这种表达能够描述任意形态物体,抑制杂乱背景和遮挡,几何意义直观,且适应不同分辨率、非规则的三角格网.然后,以这种表达方法为基础,本文阐述了一种基于多分辨率格网的,由粗到精的三维物体识别方法.它先在场景和模型的低分辨率格网上进行粗匹配以得到模型候选集合,之后在已匹配网格的高分辨率格网邻域上筛选模型候选集合,最后综合考虑多个网格对应的模型候选以得到最终模型候选的确认和验证.这种识别方法具有运算量小,准确可靠等优点,实验证明该方法正确有效.     

Optimized quadtree for Karhunen-Loeve transform in multispectralimage coding

A new multispectral image compression technique based on the Karhunen-Loeve transform (KLT) and the discrete cosine transform (DCT) is proposed. The quadtree for determining the transform block size and the quantizer for encoding the transform coefficients are jointly optimized in a rate-distortion sense. The problem is solved by a Lagrange multiplier approach. After a quadtree is determined by this approach, a one-dimensional (1-D) KLT is applied to the spectral axis for each block before the DCT is applied on the spatial domain. The eigenvectors of the autocovariance matrix, the quantization scale, and the quantized transform coefficients for each block are the output of the encoder. The overhead information required in this scheme is the bits for the quadtree, KLT, and quantizer representation.     

圆柱面上的轴向偶极子阵的低副瓣方向图优化综合

本文利用新的非线性最优化方法讨论圆柱面上的轴向偶极子阵的低副瓣方向图优化综合问题，利用阵列的单元可实现增益方向图，我们把阵列综合问题表述为一个约束非线性规划问题，采用新直接法求解该问题即可得出一组使阵列的总方向图满足副瓣指标要求，同时具有尽可能高的方向性系数的激励系数，如果阵列的总方向图不能达到副瓣指标，则可得出一组使阵列方向图在可达到的副瓣方向图综合设计的模拟计算结果。     

The simultaneous interpolation of antenna radiation patterns inboth the spatial and frequency domains using model-based parameterestimation

The Pade rational function fitting model commonly used for model-based parameter estimation (MBPE) in the frequency domain is enhanced to include spatial dependence in the numerator and denominator coefficients. This allows the function to interpolate an antenna radiated electric field pattern in both the frequency and spatial domains simultaneously, such that a single set of coefficients can be used to accurately reconstruct an entire radiation pattern at any frequency in the fitting-model range. A simple procedure is introduced for transforming interpolated electric fields into gain patterns using input impedance versus frequency curves also obtained via MBPE. The utility of this method is demonstrated by applying it to a dipole antenna over a frequency range of 150-950 MHz and using a polynomial representation in &thetas; for the coefficient spatial dependence. It is also used to estimate radiation patterns for a three-element Yagi array between the frequencies of 470 and 500 MHz using a binomial representation for the spatial variation that includes terms dependent on &thetas; as well as φ. The use of this method for interpolating radiation patterns has at least two significant advantages; one being large compression ratios for the amount of data that must be stored to accurately reproduce patterns and the other being a significant decrease in the amount of time required for modeling problems with large computational domains     

Multiple Emitters Localization by UAV with Nested Linear Array: System Scheme and 2D-DOA Estimation Algorithm

Unmanned Aerial Vehicle(UAV)equipped with uniform linear array has been applied to multiple emitters localization.Meanwhile,nested linear array enables to enhance localization resolution and achieve under-determined Direction of Arrival(DOA)estimation.In this paper,we propose a new system structure for emitters localization that combines the UAV with nested linear array,which is capable of significantly increasing the positioning accuracy of interested targets.Specifically,a localization scheme is designed to obtain the paired two-dimensional DOA(2D-DOA,i.e.azimuth and elevation angles)estimates of emitters by nested linear array with UAV.Furthermore,we propose an improved DOA estimation algorithm for emitters localization that utilizes Discrete Fourier Transform(DFT)method to obtain coarse DOA estimates,subsequently,achieve the fine DOA estimates by sparse representation.The proposed algorithm has lower computational complexity because the coarse DOA estimates enable to shrink the range of over-complete dictionary of sparse representation.In addition,compared to traditional uniform linear array,improved 2D-DOA estimation performance of emitters can be obtained with a nested linear array.Extensive simulation results testify the effectiveness of the proposed method.     

Multi-pose 3D face recognition based on 2D sparse representation

The increasing availability of 3D facial data offers the potential to overcome the difficulties inherent with 2D face recognition, including the sensitivity to illumination conditions and head pose variations. In spite of their rapid development, many 3D face recognition algorithms in the literature still suffer from the intrinsic complexity in representing and processing 3D facial data. In this paper, we propose the intrinsic 3D facial sparse representation (I3DFSR) algorithm for multi-pose 3D face recognition. In this algorithm, each 3D facial surface is first mapped homeomorphically onto a 2D lattice, where the value at each site is the depth of the corresponding vertex on the 3D surface. Each 2D lattice is then interpolated and converted into a 2D facial attribute image. Next, the sparse representation is applied to those attribute images. Finally, the identity of each query face can be obtained by using the corresponding sparse coefficients. The innovation of our approach lies in the strategy of converting irregular 3D facial surfaces into regular 2D attribute images such that 3D face recognition problem can be solved by using the sparse representation of those attribute images. We compare the proposed algorithm to three widely used 3D face recognition algorithms in the GavabDB database, to six state-of-the-art algorithms in the FRGC2.0 database, and to three baseline algorithms in the NPU3D database. Our results show that the proposed I3DFSR algorithm can substantially improve the accuracy and efficiency of multi-pose 3D face recognition.     

一种用于分布式阵列的球谐波域声源定位方法

本文提出一种适用于任意阵型和阵元指向性的球谐波域声源定位方法,能够在较宽的频域范围内,尤其是低频,提供较高的空间分辨率。水下噪声源的高分辨识别具有重大意义。传统阵列信号处理方法对低频噪声源的精确定位要求阵列具有较大孔径和较多的阵元数,导致系统过于庞大且成本较高。我们基于声场的球谐波表达和变换,采用分布在一定空间区域内的多个阵列估计该区域的球谐波系数矢量,对系数矢量进行信号处理实现声源定位。理论证明了该方法具有理想的空间选择特性。在一种特定的阵元分布下,仿真研究了该算法的方位谱估计性能以及阵元不一致性和位置误差对声源定位性能的影响。仿真结果显示,该算法在低频具有较高的空间分辨率且误差对算法定位性能的影响有限。      

A new low-side-lobe pattern synthesis technique for conformalarrays

The application of the new nonlinear optimization algorithms to array design is demonstrated by the low-side-lobe pattern synthesis of conformal arrays. By adopting the information of array element realized gain patterns the authors formulate synthesis problems as nonlinearly constrained optimization problems, and propose a direct method to solve them. The technique allows them to find a set of array coefficients that yield a pattern meeting a specified side-lobe level and achieving the maximum directivity, if such a set of array coefficients exists. If the side-lobe specifications cannot be met with the given array, the technique will result in a set of coefficients that yield a pattern meeting the best attainable side-lobe level and having directivity as high as possible. Simplified synthesis problems for an axial dipole array in an infinite, perfectly conducting cylinder are discussed, and numerical results are given. The synthesis technique works for general array patterns     

Mathematical representation of 2D image boundary contour using fractional implicit polynomial

Implicit polynomial (IP) fitting is an effective method to quickly represent two-dimensional (2D) image boundary contour in the form of mathematical function. Under the same maximum degree, the fractional implicit polynomial (FIP) can express more curve details than IP and has obvious advantages for the representation of complex boundary contours. In existing studies, algebraic distance is mainly used as the fitting objective of the polynomial. Although the time cost is reduced, there are problems of low fitting accuracy and spurious zero set. In this paper, we propose a two-stage neural network with differentiable geometric distance, which uses FIP to achieve mathematical representation, called TSEncoder. In the first stage, the continuity constraint is used to obtain a rough outline of the fitting target. In the second stage, differentiable geometric distance is gradually added to fine-tune the polynomial coefficients to obtain a contour representation with higher accuracy. Experimental results show that TSEncoder can achieve mathematical representation of 2D image boundary contour with high accuracy.     

Signed-digit-based multiplier-free realizations for multirateconverters

We propose multiplier-free realizations for FIR multirate converters. Radix-r signed-digit number representation and periodically time-varying (PTV) coefficients are used to obtain multiplier-free realizations. The PTV coefficients, together with one hold-and-scale unit at the input and one scale-add-downsample unit at the output, are utilized effectively to distribute the required computations to the extent that coefficients are highly simplified. The coefficients in the realizations are designed to take values from either the ternary set (0, ±1) or the quinary set (0, ±1, ±2). Their values can be easily obtained by mapping from the target filter coefficients using a radix-r signed-digit number representation. We show that fairly high precision can be achieved by these realizations     

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.     

基于多元Laplace语音模型的语音增强算法

传统的短时谱估计语音增强算法通常假设语音谱分量相互独立,没有考虑语音谱分量间的相关性。针对这一问题,该文提出一种新的基于多元Laplace分布模型的短时谱估计算法。首先,假设语音的离散余弦变换(DCT)系数服从多元Laplace分布,以此利用谱分量间的相关性;在此基础上,利用多元随机矢量的高斯尺度混合模型表示,推导得到语音DCT系数矢量的最小均方误差(MMSE)估计的解析表达式;并进一步推导了基于该分布模型的语音存在概率,对最小均方误差估计子进行修正。实验结果表明,该算法在抑制背景噪声和减少语音失真等方面优于传统的语音增强方法。     

Infinite families of quasiperfect and doubly quasiperfect binary arrays

A quasiperfect (doubly quasiperfect) binary array B is an s*t array with all entries plus or minus one, such that all periodic autocorrelation coefficients of the array are zero, except for shifts (u,v) where u identical to 0 mod s and v identical to 0 mod t. The authors construct a new infinite families of quasiperfect and of doubly quasiperfect binary arrays.<>     

基于加权小波的三维网格多描述编码

鉴于网格小波系数特点与图像小波系数的差异性,提出一种基于加权小波系数的多描述编码(WW_MDC)方法.利用小波变换,提取模型最粗糙网格和小波系数,对小波系数进行加权处理,使其更加适合随后的SPIHT算法,利用SPIHT码流分层内嵌特性将小波系数分组形成多个描述,可以很好的适应网络带宽的变化和丢包率(PLR)的变化.试验结果表明,该方法具有较高的压缩率和抗误码能力,并且相比较其他三维模型的多描述编码方法,该方法具有更小的重建误差.     

Tomographic reconstruction and estimation based on multiscalenatural-pixel bases

We use a natural pixel-type representation of an object, originally developed for incomplete data tomography problems, to construct nearly orthonormal multiscale basis functions. The nearly orthonormal behavior of the multiscale basis functions results in a system matrix, relating the input (the object coefficients) and the output (the projection data), which is extremely sparse. In addition, the coarsest scale elements of this matrix capture any ill conditioning in the system matrix arising from the geometry of the imaging system. We exploit this feature to partition the system matrix by scales and obtain a reconstruction procedure that requires inversion of only a well-conditioned and sparse matrix. This enables us to formulate a tomographic reconstruction technique from incomplete data wherein the object is reconstructed at multiple scales or resolutions. In case of noisy projection data we extend our multiscale reconstruction technique to explicitly account for noise by calculating maximum a posteriori probability (MAP) multiscale reconstruction estimates based on a certain self-similar prior on the multiscale object coefficients. The framework for multiscale reconstruction presented can find application in regularization of imaging problems where the projection data are incomplete, irregular, and noisy, and in object feature recognition directly from projection data.     

Two-Step Sparse Representation Based 2D DOA Estimation with Single Moving Acoustic Vector Sensor

This paper investigates the two dimensional direction of arrival (2D DOA) estimation problem of multiple sources with one single moving acoustic vector sensor (AVS). We first use one single moving AVS to construct a synthetic nested AVS array, which is later shown that is equivalent to the physical nested AVS array. Then the vectorization and row extraction operations are performed to obtain the observation vector that behaves like signals received by a virtual uniform AVS array. Finally, the 2D DOA estimation is obtained via a two-step sparse representation (SR) method, which transforms the 2D grid search to a computationally efficient 1D grid search. The Cramer-Rao bound comparison between the synthetic and physical nested AVS arrays shows that these two arrays are equivalent for DOA estimation. Based on the property of the nested arrays and the full utilization of the array aperture via SR, the proposed method can achieve better estimation performance than spatial smoothing methods with nested AVS arrays and methods with uniform AVS arrays. Simulations validate the effectiveness of the proposed synthetic array method.

     

Fully synthesised decimation filter for delta-sigma A/D converters

Digital decimation filters are used in delta-sigma analogue-to-digital converters to reduce the oversampled data rate to the final Nyquist rate. This paper presents the design and implementation of a fully synthesised digital decimation filter that provides a time-to-market advantage. The filter consists of a cascaded integrator-comb filter and two cascaded half-band FIR filters. A canonical signed-digit representation of the filter coefficients is used to minimise the area and to reduce the hardware complexity of the multiplication arithmetic. Coefficient multiplications are implemented by using shifters and adders. This three-stage decimation filter is fabricated by using 0.25-μm CMOS technology with an active area of 1.36 mm² and shows 4.4 mW power consumption at a clock rate of 2.8224 MHz. Experimental results show that this digital decimation filter is suitable for use in oversampled data converters and can be applied to new processes requiring a fast redesign time. This is possible because the filter does not have process-dependent ROM or RAM circuits.