In several applications it is required to reconstruct a high-resolution noise-free image from multipath frames of undersampled low-resolution noisy images. Using the aliasing relationship between undersamples and reference image, an algorithm based on weighted recursive least-squares theory developed in wavenumber domain. This efficient because interpolation noise removal are performed recursively, highly suitable for implementation via massively parallel computational architectures...

An algorithm based on spatial tessellation and approximation of each triangle patch in the Delaunay (1934) triangulation (with smoothness constraints) by a bivariate polynomial is advanced to construct high resolution (HR) quality image from set low (LR) frames. The accompanied site-insertion for update initial HR with availability more LR frames till desired attained. This algorithm, followed post filtering, suitable real-time sequence processing because fast expected (average) time...

The attainment of super resolution (SR) from a sequence degraded undersampled images could be viewed as reconstruction the high-resolution (HR) image finite set its projections on sampling lattice. This can then formulated an optimization problem whose solution is obtained by minimizing cost function. approaches adopted and their analysis to solve are crucial, acquisition scheme important in modeling degradation process. need for model accuracy undeniable SR along with design algorithm...

This article considers the problem of reconstructing a high-resolution image from multiple undersampled, shifted, degraded frames with subpixel displacement errors. leads to formulation involving periodically shift-variant system model. The maximum posteriori (MAP) estimation scheme is used subject assumption that original modeled by stationary Markov-Gaussian random field. resulting MAP expressed as complex linear matrix equation, where characterizing involves periodic block Toeplitz (BTTB)...

Properties of wavelets with finite as well infinite support are summarised to facilitate mother wavelet selection in a chosen application. The quantitative guidelines will reduce dependence on trial-and-error schemes resorted for and underscore the importance such any application interest. In wavelet-based image sequence superresolution, studied during last four years, use B-spline is justified.

Three fast algorithms have been developed to solve the problem of data association in multitarget tracking clutter. In first algorithm, is identified as an exhaustive search general. Subsequently, a mathematical model proposed for joint probabilistic filter (JPDAF). Based on model, depth-first (DFS) approach generation hypotheses and computation conditional probabilities JPDAF. When density targets moderate, second algorithm directly compute posteriori JPDAF without generating hypotheses....

An irregularly spaced sampling raster formed from a sequence of low-resolution frames is the input to an image superresolution algorithm whose output set intensity values at desired high-resolution grid. The method moving least squares (MLS) in polynomial space has proved be useful filtering noise and approximating scattered data by minimizing weighted mean-square error norm, but introducing blur process. Starting with continuous version MLS, explicit expression for filter bandwidth obtained...

A novel approach is proposed which determines the number of layers, neurons in each layer, and their connection weights for a particular implementation neural network, with multilayer feedforward topology, designed to classify patterns multidimensional feature space. The based on construction Voronoi diagram over set points representing space this finds added usefulness deriving alternate network structures realizing desired pattern classification.< <ETX...

First, a sufficiency test is developed to for all the zeros of polynomial be within unit circle when each its coefficients takes an arbitrary but fixed value specified compact interval. This robust stability discrete-time systems implemented via use bilinear transformation and Kharitonov&apos;s theorem. Subsequently, this result used set up novel 2-dimensional digital filters. It shown that last can generalised tests n-dimensional (n &gt; 2) An example constructed illustrate principles...

It is shown how the total least squares recursive algorithm for real data FIR (finite impulse response) adaptive filtering problem can be applied to reconstruct a high-resolution filtered image from undersampled, noisy multiframes, when interframe displacements are not accurately known. This done in wavenumber domain after transforming complex an equivalent problem, which developed by C.E. Davila (Proc. ICASSP 1991 p.1853-6 of 1991) applies. The procedure also applies multiframes degraded...

Morf, Levy, and Kung Youla Gnavi presented a primitive factorization algorithm which extracts in some sense the content of (full rank) matrix <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</tex> with entries ring xmlns:xlink="http://www.w3.org/1999/xlink">K[z,\omega]</tex> bivariate polynomials over field xmlns:xlink="http://www.w3.org/1999/xlink">K</tex> . However, algorithms both cases specify require coefficient to be algebraically...

An algorithm with a view towards computer implementation is given for stability test of filters arbitrary dimension and complexity. The based on the generation number multivariable polynomials, reduction each these into several single variable polynomials by finite rational operations, scheme repeated polynomial factorizations back substitutions.

In this paper, a systematic procedure to test for stability of three-dimensional filters (discrete and continuous) is presented. The based on repeated applications an extended Hermite or Schur-Cohn formulation, use Sturm's theorem determine the content system polynomial inequalities in single indeterminate. need generating constructive algorithm tests higher than using Tarski's generalization discussed. Application certain combinatorial rules transforming multidimensional digital filter...

Abstract Multiple undersampled images of a scene are often obtained by using charge‐coupled device (CCD) detector array sensors that shifted relative to each other subpixel displacements. This geometry sensors, where sensor has subarray sensing elements suitable size, been popular in the task attaining spatial resolution enhancement from acquired low‐resolution degraded comprise set observations. With objective improving performance signal processing algorithms presence ubiquitous...

This paper is a tutorial on Grobner bases and survey the applications of in broad field signals systems. A reasonably detailed review given several fundamental theoretical issues that occur use multidimensional systems applications. These topics include primeness multivariate polynomial matrices, unimodular matrix completion, prime factorization matrices. brief also presented wide-ranging as well one-dimensional circuits, networks, control, coding, signals, other related areas like robotics...

In view of several potential applications in multivariable two-dimensional (2-D) systems theory, the questions existence and nonuniqueness solutions along with scopes for recursive computation when is guaranteed, are investigated. Attention directed towards exploitation a block Hankel-Hankel matrix structure that can be used to characterize problem. Some known results on 1-D 2-D scalar cases as well very recently presented case result fall out nicely special research reported here.

Abstract By introducing a multivariable form of Tellegen's Theorem and generalization energy functions, the positive-real nature immittance matrices finite networks composed resistors, p-plane inductors capacitors, all non-negative, transformers gyrators, is proven, us well as lossless constraint. Multivariable reactance functions are investigated relation between such Hurwitz polynomials proved.

A recently introduced new stability test for single-dimensional digital filters is extended to apply two-dimensional filters, and the procedure implementation of this discussed. Nontrivial examples both discrete continuous are used illustrate procedure.

A multivariate polynomial matrix factorization algorithm is introduced and discussed. This another for computing a globally minimal generating of the syzygy solutions associated with are both zero-coprimeness constraint that characterizes perfect-reconstruction filter banks. Generalizations, as well limitations recent results which incorporate perfect reconstruction linear-phase constraints, discussed several examples counterexamples. Specifically, Grobner basis-based proof linear phase...

Subsequent to the work of Kim, Bose, and Valenzuela in 1990 on simultaneous filtering interpolation a registered sequence undersampled noisy shift-invariant blur degraded images, Bose Boo tackled 1998 problem reconstructing high-resolution image from multiple undersampled, shifted, frames with subpixel displacement errors. This led formulation involving periodically shift-variant system model. Lertrattanapanich advanced 1999 procedure for construction video mosaic following estimation motion...