Perceptual primitives from an extended 4D Hough transform

Directional features extracted from Gabor wavelets responses were used to train a structure of self-organising maps, thus classifying each pixel in the image within a neuron-map. Resulting directional primitives were grouped into perceptual primitives introducing an extended 4D Hough transform to group pixels with similar directional features. These can then be used as perceptual primitives to detect salient structures. The proposed method has independently fixed parameters that do not need to be tuned for different kind or quality of images. We present results in application to noisy FLIR images and show that line primitives for complex structures, such as bridges, or simple structures, such as runways, can be found by this approach. We compare and demonstrate the quality of our results with those obtained through a parameter-dependent traditional Canny edge detector and Hough line finding process.     

Wavelet packet based channel equalization

Recently, considerable amount of attention is being given to the field of wavelets and wavelet packets. It has found numerous applications in signal representation, image compression and applied mathematics. In this paper, we present a channel equalization method based on wavelet packets. The proposed equalizer structure is based on the fact that for sufficiently narrowband sequences, a non-ideal channel can be modelled as an attenuation and delay. If the data sequence is used to modulate a set of narrowband wavelet packets, then no equalization is required at the receiver end. The equalization problem reduces to that of determining the delay introduced by the channel for each of the wavelet packets. A minimum square variance algorithm for adaptively choosing the delay has been proposed. This algorithm has been shown to perform as desired analytically in a simple delay channel case. Simulations have been used to study its performance in the non-ideal channel’s case and the results corroborate theoretical predictions.     

A Review of Wavelets for Digital Wireless Communication

Wavelets have been favorably applied in almost all aspects of digital wireless communication systems including data compression, source and channel coding, signal denoising, channel modeling and design of transceivers. The main property of wavelets in these applications is in their flexibility and ability to characterize signals accurately. In this paper recent trends and developments in the use of wavelets in wireless communications are reviewed. Major applications of wavelets in wireless channel modeling, interference mitigation, denoising, OFDM modulation, multiple access, Ultra Wideband communications, cognitive radio and wireless networks are surveyed. The confluence of information and communication technologies and the possibility of ubiquitous connectivity have posed a challenge to developing technologies and architectures capable of handling large volumes of data under severe resource constraints such as power and bandwidth. Wavelets are uniquely qualified to address this challenge. The flexibility and adaptation provided by wavelets have made wavelet technology a strong candidate for future wireless communication. Madan Kumar Lakshmanan was born in Chennai, India, in 1979. He received the B.E. (with distinction) in electrical engineering from the University of Madras, Chennai, India, in 2000. He joined the Indian Software firm, Polaris Software Labs Ltd., in 2000 where he wrote software for Telecommunication applications. At Polaris, he was awarded the “On The Spot Of Excellence Award” for his efforts. In 2003, he moved to the Indian Institute of Technology-Madras, to develop and establish a wireless communications network for rural connectivity. In 2004, he was awarded the Royal Dutch/Shell Chevning scholarship to pursue a Master degree in Telecommunications at the Delft University of Technology (TUDelft). At TUDelft he is affiliated to the International Research Center for Telecommunications-Transmission and Radar (IRCTR) where he is undertaking research in the field of wavelets applications in Wireless Communications. Homayoun Nikookar received his Ph.D. in Electrical Engineering from Delft University of Technology (TUDelft), The Netherlands, in 1995. From 1995 to 1998 he was a postdoc researcher at the International Research Center for Telecommunications-Transmission and Radar, TUDelft, where since 1999 he has been an Assistant Professor. Dr. Nikookar has done research on different areas of wireless communications, including wireless channel modeling, UWB, MIMO, multicarrier transmission, Wavelet-based OFDM and CDMA. He is a senior member of the IEEE.     

Nonlinear black-box modeling in system identification: a unified overview

A nonlinear black-box structure for a dynamical system is a model structure that is prepared to describe virtually any nonlinear dynamics. There has been considerable recent interest in this area, with structures based on neural networks, radial basis networks, wavelet networks and hinging hyperplanes, as well as wavelet-transform-based methods and models based on fuzzy sets and fuzzy rules. This paper describes all these approaches in a common framework, from a user's perspective. It focuses on what are the common features in the different approaches, the choices that have to be made and what considerations are relevant for a successful system-identification application of these techniques. It is pointed out that the nonlinear structures can be seen as a concatenation of a mapping form observed data to a regression vector and a nonlinear mapping from the regressor space to the output space. These mappings are discussed separately. The latter mapping is usually formed as a basis function expansion. The basis functions are typically formed from one simple scalar function, which is modified in terms of scale and location. The expansion from the scalar argument to the regressor space is achieved by a radial- or a ridge-type approach. Basic techniques for estimating the parameters in the structures are criterion minimization, as well as two-step procedures, where first the relevant basis functions are determined, using data, and then a linear least-squares step to determine the coordinates of the function approximation. A particular problem is to deal with the large number of potentially necessary parameters. This is handled by making the number of ‘used’ parameters considerably less than the number of ‘offered’ parameters, by regularization, shrinking, pruning or regressor selection.     

一种基于双正交重叠变换的嵌入式编码算法研究

双正交变换为基于块的低复杂度变换,而且与传统的离散余弦变换相比,一定程度上减小了变换后图像的块效应,因而被采纳入联合图像专家组织JPEG（Joint Photographic Expert Group）最新制定的静态图像编码标准JPEG XR中。为了能够改善其无法实现码流长度控制的缺陷,文章深入研究了JPEG XR的编码技术,提出了一种针对固定压缩比的编码算法。主要思路是通过对双正交变换后的系数进行嵌入式位平面编码,取代了原先的量化步骤,使得压缩码流长度可以精确控制。     

A composite numerical scheme for the numerical simulation of coupled Burgers’ equation

In this work, a composite numerical scheme based on finite difference and Haar wavelets is proposed to solve time dependent coupled Burgers’ equation with appropriate initial and boundary conditions. Time derivative is discretized by forward difference and then quasilinearization technique is used to linearize the coupled Burgers’ equation. Space derivatives discretization with Haar wavelets leads to a system of linear equations and is solved using Matlab7.0. Convergence analysis of proposed scheme exhibits that the error bound is inversely proportional to the resolution level of the Haar wavelet. Finally, the adaptability of proposed scheme is demonstrated by numerical experiments and shows that the present composite scheme offers better accuracy in comparison with other existing numerical methods.     

具有样条对偶的散度自由小波

选取2对在弱对偶意义下满足微分关系的B-样条作为尺度函数,构造了L2(n)n上具有样条对偶的散度自由单小波,证明了不可压缩流体向量场在弱对偶意义下的投影保持散度自由性质,并给出了具体的例子.     

Gabor小波和改进LBP的零件表面粗糙度识别

目的 为了提取零件表面图像的纹理特征并对其表面粗糙度分类识别,有效提高识别的正确率,提出了联合Gabor小波和改进局部二值模式(LBP)的纹理特征提取方法。方法 针对传统LBP算子忽略了邻域内灰度差幅值特征的问题,提出了M_LBP(magnitude considered LBP)算子。采用Gabor小波对零件表面图像滤波,并计算各子图像 Gabor幅值特征GMM(Gabor magnitude maps)。应用M_LBP算子计算各GMM的M_LBP特征谱,进而构造得到零件表面图像的纹理特征向量,最后通过KNN(K-nearest neighbor)算法对零件粗糙度分类识别。结果 本文提出的算法有效细化了表面图像纹理特征,对粗糙度差别为0.2 μm的零件识别准确率达到98%,远高于利用传统LBP算子提取的纹理信息的识别准确率。结论 本文提出了一种有效细化LBP纹理特征的M_LBP算子,并通过与Gabor小波的结合,突破了传统LBP算子尺度、方向单一,幅值信息被忽略的局限性,能实现较高精度的粗糙度识别。     

新 Burt-Adelson双正交小波的优化设计 *

根据小波滤波器的精确重构条件 ,推导出 Burt-Adelson双正交小波滤波器组的参数表达式。该表达式参数可任意取值 ,因此能够随意构造具有不同特征的 Burt-Adelson小波。作为构造实例 ,构造出一个新的有理系数 Burt-Adelson小波 ,它具有优化的编码增益。实验表明 ,其图像压缩性能略低于 CDF-9 /7小波 ,但优于 JPEG 2000标准推荐的 7 /5小波 ;而且其提升小波变换的计算效率比 CDF-9 /7小波高 20%以上 ,也优于 7 /5小波。     

CAS wavelet method for solving the fractional integro-differential equation with a weakly singular kernel

In this paper, a computational method for numerical solution of a class of integro-differential equations with a weakly singular kernel of fractional order which is based on Cos and Sin (CAS) wavelets and block pulse functions is introduced. Approximation of the arbitrary order weakly singular integral is also obtained. The fractional integro-differential equations with weakly singular kernel are transformed into a system of algebraic equations by using the operational matrix of fractional integration of CAS wavelets. The error analysis of CAS wavelets is given. Finally, the results of some numerical examples support the validity and applicability of the approach.