Homogeneous Random Field Research Articles

An Edge Preserving Multiresolution Fusion: Use of Contourlet Transform and MRF Prior

In this paper, we propose a new approach for multiresolution fusion using contourlet transform (CT). The method is based on modeling the low spatial resolution (LR) and high spectral resolution multispectral (MS) image as the degraded and noisy version of their high spatial resolution version. Since this is an ill-posed problem, it requires regularization in order to obtain the final solution. In this paper, we first obtain the initial estimate of the fused image from the available MS image and the panchromatic (Pan) image by using the CT domain learning. Since CT provides better directional edges, the initial estimate has better edge details. Using the initial estimate, we obtain the degradation that accounts for the aliasing between the LR MS image and fused image. Regularization is carried out by modeling the texture of the final fused image as a homogeneous Markov random field (MRF) prior, where the MRF parameter is estimated using the initial estimate. The use of MRF prior on the final fused image takes care of the spatial dependencies among the pixels. A simple gradient-based optimization technique is used to obtain the final fused image. Although we use homogeneous MRF, the proposed approach preserves the edges in the final fused image by retaining the edges from the initial estimate and by carrying out the optimization on nonedge pixels only. Therefore, the advantage of the proposed method lies in preserving the discontinuities without using the discontinuity preserving prior, thus avoiding the use of computationally taxing optimization techniques for regularization purposes. In addition, the proposed method causes minimum spectral distortion since it learns the texture using contourlet coefficients and does not use actual Pan image pixel intensities. We demonstrate the effectiveness of our approach by conducting the experiments using subsampled and nonsubsampled CT on different data sets captured using Ikonos-2, Quickbird, and Worldview-2 satellites.

Characterization of random stress fields obtained from polycrystalline aggregate calculations using multi-scale stochastic finite elements

The spatial variability of stress fields resulting from polycrystalline aggregate calculations involving random grain geometry and crystal orientations is investigated. A periodogram-based method is proposed to identify the properties of homogeneous Gaussian random fields (power spectral density and related covariance structure). Based on a set of finite element polycrystalline aggregate calculations the properties of the maximal principal stress field are identified. Two cases are considered, using either a fixed or random grain geometry. The stability of the method w.r.t the number of samples and the load level (up to 3.5% macroscopic deformation) is investigated.

On Piterbarg's max-discretisation theorem for homogeneous Gaussian random fields

Motivated by the papers of Piterbarg [17] and Hüsler [7], in this paper the asymptotic relation between the maximum of a continuous dependent homogeneous Gaussian random field and the maximum of this field sampled at discrete time points is studied. It is shown that, for the weakly dependent case, these two maxima are asymptotically independent, dependent or coincide when the grid of the discrete time points is a sparse grid, Pickands grid or dense grid, respectively, while for the strongly dependent case, these two maxima are asymptotically totally dependent if the grid of the discrete time points is sufficiently dense, and asymptotically dependent if the grid points are sparse or Pickands grids.

An efficient coupling of FORM and Karhunen–Loève series expansion

The topic of this paper is the solution of reliability problems where failure is influenced by the spatial random fluctuations of loads and material properties. Homogeneous random fields are used t...

Lattice vector quantisation for indexing and retrieval of medical images using texture features based on 2-D Wold decomposition

Medical image retrieval to search for clinically relevant and visually similar images to support radiologists' examination has been attracting lot of research interest. It provides a baseline essential for confirmation, comparison and evaluation of suspicious radiographic signs detected on current examination images. Content-based image retrieval (CBIR) is an important alternate and complement to traditional text-based image retrieval (TBIR) using keywords. The proposed CBIR system is based on the effective use of texture information within images using 2-D Wold decomposition of homogeneous random fields. To speed up the search process, selected features are extracted and indexed using the proposed indexing structure. This paper describes a new efficient indexing structure for elongated vectors of feature space using lattice vector quantisation (LVQ). Selected features are indexed based on norm, leader and permutation indices. Simple Euclidean distance measure is used to display the retrieved images in the decreasing order of similarity. Features for harmonic structures and evanescent components are compared based on precision and recall. Experiments were carried out on MR-T2 axial brain slices. Features of harmonic structure gives better results when compared to evanescent components.

Searching regions of interest on vector local homogeneous scenes

A generalized solution of the problem of searching a scene's fragments, called the regions of interest, for the case when a set of spatial superposed images serves as the initial information about the scene, is proposed. A vector homogeneous random field is taken to be the scene model.

Numerical simulation and inversion validation of the real correlation random wind field generation for coherent Doppler LiDAR

Simulation of the coherent Doppler LiDAR signal requires accurate computation of homogeneous random wind fields. Based on complex random processes with specified spatial statistics given by the covariance function, an improved real correlation random wind field algorithm is proposed for real random processes, the simulation results are compared with the given covariance function and the real correlation algorithm conforms to the given covariance function quite well.Moreover, to validate the accuracy of the real correlation algorithm, the estimation of the turbulence energy dissipation rate (TEDR) from the Doppler spectrum width is described. Using the generated real random wind field, the estimation relative bias of the TEDR is retrieved. The accuracy of the simulation is in accordance with the experimental results.

A stochastic model for earthquake slip distribution of large events

This paper presents a stochastic model to simulate spatial distribution of slip on the rupture plane for large earthquakes (Mw >7). A total of 45 slip models coming from the past 33 large events are examined to develop the model. The model has been developed in two stages. In the first stage, effective rupture dimensions are derived from the data. Empirical relations to predict the rupture dimensions, mean and standard deviation of the slip, the size of asperities and their location from the hypocentre from the seismic moment are developed. In the second stage, the slip is modelled as a homogeneous random field. Important properties of the slip field such as correlation length have been estimated for the slip models. The developed model can be used to simulate ground motion for large events.

The evolution of the nanoscale dissipative structures in a distribution of defects within the isothermally irradiated f.c.c. crystal

A kinetic model for the influence of external noises such as fluctuations of the vacancies’ generation rate and inhomogeneity of irradiated f.c.c. crystal on the formation of nanoscale modulated dissipative structure in a spatial distribution of vacancies is considered. The generation rate of vacancies all over the sites and a density of their dislocation-type sinks are modelled as independent random uniform stationary fields and with certain defined parameters of fluctuation correlations – spatial and temporal ones. Such stochastic fields can induce a spatial redistribution of vacancies that can lead to their density stationary uniform field or stochastic one. By the average value and correlation functions of these fluctuations, the conditions are determined for interacting fluctuations of the vacancies’ density, under which this homogeneous random field becomes unstable in relation to the stochastic field with a spatially periodic mean distribution of vacancies’ density. For instance, with f.c.c. nickel as a model of the irradiated functional material, the temperature dependence of spatial period d(T) of the modulated dissipative structure of vacancies’ subsystem in f.c.c. crystal is numerically forecasted and analysed, taking into account the total (‘electrochemical’ + ‘strain-induced’) interaction between vacancies. Such d(T)-dependence is also determined by the kinetic characteristics of vacancies’ redistribution.

The Sequential Generation of Gaussian Random Fields for Applications in the Geospatial Sciences

This paper presents practical methods for the sequential generation or simulation of a Gaussian two-dimensional random field. The specific realizations typically correspond to geospatial errors or perturbations over a horizontal plane or grid. The errors are either scalar, such as vertical errors, or multivariate, such as , , and errors. These realizations enable simulation-based performance assessment and tuning of various geospatial applications. Both homogeneous and non-homogeneous random fields are addressed. The sequential generation is very fast and compared to methods based on Cholesky decomposition of an a priori covariance matrix and Sequential Gaussian Simulation. The multi-grid point covariance matrix is also developed for all the above random fields, essential for the optimal performance of many geospatial applications ingesting data with these types of errors.

Statistically isotropic tensor random fields: Correlation structures

Let V be a real finite-dimensional vector space. We introduce some physical problems that may be described by V-valued homogeneous and isotropic random fields on R 3 . We propose a general method f ...

Matrix Crack Detection in Composite Plate with Spatially Random Material Properties using Fractal Dimension

Fractal dimension based damage detection method is investigated for a composite plate with random material properties. Composite material shows spatially varying random material properties because of complex manufacturing processes. Matrix cracks are considered as damage in the composite plate. Such cracks are often seen as the initial damage mechanism in composites under fatigue loading and also occur due to low velocity impact. Static deflection of the cantilevered composite plate with uniform loading is calculated using the finite element method. Damage detection is carried out based on sliding window fractal dimension operator using the static deflection. Two dimensional homogeneous Gaussian random field is generated using Karhunen-Loeve (KL) expansion to represent the spatial variation of composite material property. The robustness of fractal dimension based damage detection method is demonstrated considering the composite material properties as a two dimensional random field.

Generalized variability response functions for two-dimensional elasticity problems

The ability to determine probabilistic characteristics of response quantities in structural mechanics (e.g. displacements, stresses) as well as effective material properties is restricted due to lack of information on the probabilistic characteristics of the uncertain system parameters. The concept of the variability response function (VRF) has been proposed as a means to systematically capture the effect of the stochastic spectral characteristics of uncertain system parameters modeled by homogeneous random fields on the uncertain structural response. The key property of the VRF in its classical sense is its independence from the marginal probability distribution function (PDF) and the spectral density function (SDF) of the uncertain system parameters (it depends only on the deterministic structural configuration and boundary conditions). Proofs have been provided for the existence of VRFs for linear and some nonlinear statically determinate beams. For statically indeterminate structures, the Monte Carlo based generalized variability response function (GVRF) methodology has been proposed recently as a generalization of the VRF concept to indeterminate linear and some nonlinear beams. The methodology computes GVRFs, which are analogous to VRFs for statically determinate structures, and evaluates their dependence (or lack thereof) on the PDF and SDF of the random field, thereby providing an estimate of the accuracy of the GVRF. In this paper, the GVRF methodology is extended to problems involving two-dimensional, linear continua whose stochasticity is characterized by statistically homogeneous random fields. After detailing the GVRF methodology for two-dimensional random fields, two numerical examples are provided: GVRFs are computed for the displacement response and for the effective compliance of linear plane stress systems.

Anisotropic statistical scaling of vadose zone hydraulic property estimates near Maricopa, Arizona

[1] Fluid flow and mass transport in the vadose zone are strongly influenced by spatial variability in soil hydraulic properties. It has become common to characterize spatial variability geostatistically and to solve corresponding flow and/or transport problems stochastically. The typical (though not only) approach is to treat log saturated hydraulic conductivity, Y = log10 Ks, and perhaps other medium properties as statistically homogeneous, isotropic, or anisotropic multivariate-Gaussian random fields with unique variance and autocorrelation scale. A growing body of literature suggests that Y as well as many other variables may possess heavy-tailed, non-Gaussian distributions, and/or scale-dependent statistical parameters representing a multiscale, hierarchical structure. Elsewhere, we have demonstrated that these important phenomena are difficult to detect, and are not fully tractable, with standard geostatistical methods. They are however detectable and tractable with a novel geostatistical method of analysis which treats such variables as samples from sub-Gaussian random fields subordinated to truncated fractional Brownian motion (tfBm) or truncated fractional Gaussian noise (tfGn). Each such field is a mixture of Gaussian components having random variances. The purpose of this paper is to explore the extent to which hydraulic parameters of unsaturated soils exhibit heavy-tailed distributions and statistical scaling of the above type. As a test bed we have selected an experimental site near Maricopa, Arizona, at which soil texture data have been collected in boreholes and trenches to a depth of 15 m over an area of 3600 m2. Since hydraulic parameters are difficult to measure to such depths we estimate them using a neural network pedotransfer model. Input data include percent sand, silt, and clay. Outputs include estimates of saturated and residual volumetric water content, saturated hydraulic conductivity, and (due to their wide applicability) parameters of the van Genuchten-Mualem constitutive model of unsaturated soil property variations with capillary pressure head. We find indeed that vertical and horizontal increments of our hydraulic parameter estimates exhibit heavy-tailed frequency distributions and statistical scaling which conform closely to our novel geostatistical framework. Its generality and applicability to the Maricopa site has far reaching implications vis-à-vis the geostatistical characterization of vadose zone properties and the stochastic modeling of flow and transport in unsaturated media at other sites.

Strictly and non-strictly positive definite functions on spheres

Isotropic positive definite functions on spheres play important roles in spatial statistics, where they occur as the correlation functions of homogeneous random fields and star-shaped random particles. In approximation theory, strictly positive definite functions serve as radial basis functions for interpolating scattered data on spherical domains. We review characterizations of positive definite functions on spheres in terms of Gegenbauer expansions and apply them to dimension walks, where monotonicity properties of the Gegenbauer coefficients guarantee positive definiteness in higher dimensions. Subject to a natural support condition, isotropic positive definite functions on the Euclidean space $\mathbb{R}^3$, such as Askey's and Wendland's functions, allow for the direct substitution of the Euclidean distance by the great circle distance on a one-, two- or three-dimensional sphere, as opposed to the traditional approach, where the distances are transformed into each other. Completely monotone functions are positive definite on spheres of any dimension and provide rich parametric classes of such functions, including members of the powered exponential, Mat\'{e}rn, generalized Cauchy and Dagum families. The sine power family permits a continuous parameterization of the roughness of the sample paths of a Gaussian process. A collection of research problems provides challenges for future work in mathematical analysis, probability theory and spatial statistics.

An approach to detection of point sources in very high-resolution microwave maps

This paper deals with the detection problem of extragalactic point-sources in multi-frequency, microwave sky maps that will be obtainable in future cosmic microwave background radiation (CMB) experiments with instruments capable of very high spatial resolution. With spatial resolutions that can be of order of 0.1-1.0 arcsec or better, the extragalactic point-sources will appear isolated. The same holds also for the compact structures due to the Sunyaev-Zeldovich (SZ) effect (both thermal and kinetic). This situation is different from the maps obtainable with instruments as WMAP or PLANCK where, because of the smaller spatial resolution (approximately 5-30 arcmin), the point-sources and the compact structures due to the SZ effect form a uniform noisy background (the "confusion noise"). Hence, the point-source detection techniques developed in the past are based on the assumption that all the emissions that contribute to the microwave background can be modeled with homogeneous and isotropic (often Gaussian) random fields and make use of the corresponding spatial power-spectra. In the case of very high resolution observations such an assumption cannot be adopted since it still holds only for the CMB. Here, we propose an approach based on the assumption that the diffuse emissions that contribute to the microwave background can be locally approximated by two-dimensional low order polynomials. In particular, two sets of numerical techniques are presented containing two different algorithms each. The performance of the algorithms is tested with numerical experiments that mimic the physical scenario expected for high Galactic latitude observations with the Atacama Large Millimeter/Submillimeter Array (ALMA).

RÉNYI FUNCTION FOR MULTIFRACTAL RANDOM FIELDS

This paper presents the basic scheme and the log-normal, log-gamma and log-negative-inverted-gamma scenarios to establish the Rényi function for infinite products of homogeneous isotropic random fields on Rn; in particular for random fields on the sphere in R3. The motivation of this paper is the test of (non-)Gaussianity on the Cosmic Microwave Background Radiation (CMBR) in Cosmology. In the presentation, we need to employ spherical harmonics for some concrete computations.

Macroscaling Limit Theorems for Filtered Spatiotemporal Random Fields

This article addresses the problem of defining a general scaling setting in which Gaussian and non-Gaussian limit distributions of linear random fields can be obtained. The linear random fields considered are defined by the convolution of a Green kernel, satisfying suitable scaling conditions, with a non-linear transformation of a Gaussian centered homogeneous random field. The results derived cover the weak-dependence and strong-dependence cases for such Gaussian random fields. Extension to more general random initial conditions defined, for example, in terms of non-linear transformations of χ2-random fields, is also discussed. For an example, we consider the random fractional diffusion equation. The vectorial version of the limit theorems derived is also formulated, including the limit distribution of the parabolically rescaled solution to the Burgers equation in the cases of weakly and strongly dependent initial potentials.

Asymptotic properties of the method of observed mean for homogeneous random fields

The empirical estimate of the unknown parameter of a homogenous (in restricted sense) random field with continuous time and continuous states observed in a circle is considered. The strong consistency of the estimate is proved. The conditions under which the estimates weakly converge to the Gaussian distribution are established.

Randomized Spectral and Fourier-Wavelet Methods for multidimensional Gaussian random vector fields

We develop some new variations of the Randomized Spectral Method (RSM) and Fourier-Wavelet Method (FWM) for the simulation of homogeneous Gaussian random vector fields based on plane wave decompositions. We then compare the various random field simulation methods through the examination of their efficiency and the quality of the Eulerian and Lagrangian statistical characteristics of a three-dimensional isotropic incompressible random field as simulated by ensemble or by spatial averaging. We find that the RSM, which is easier to implement, can simulate random fields with accurate second order Eulerian and Lagrangian correlations more efficiently than the FWM when the statistics are collected through ensemble averages. When the correlation structure is inferred from spatial averages of a single realization of the random field (through appeal to an ergodic theorem), the FWM is more efficient in simulating an accurate Eulerian correlation function, and both the FWM and RSM are of comparable efficiency in simulating second order Lagrangian statistics accurately.