Finite Differences in Forward and Inverse Imaging Problems: MaxPol Design
From MaRDI portal
boundary conditionnumerical differentiationderivative matriximage stitchingfullband/lowpass FIR differentiationgradient surface recoveryinverse Vandermondemaxflat technique
Eigenvalues, singular values, and eigenvectors (15A18) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10) Numerical methods for inverse problems for boundary value problems involving PDEs (65N21) Finite difference methods for boundary value problems involving PDEs (65N06) Numerical differentiation (65D25)
Abstract: A systematic and comprehensive framework for finite impulse response (FIR) lowpass/fullband derivative kernels is introduced in this paper. Closed form solutions of a number of derivative filters are obtained using the maximally flat technique to regulate the Fourier response of undetermined coefficients. The framework includes arbitrary parameter control methods that afford solutions for numerous differential orders, variable polynomial accuracy, centralized/staggered schemes, and arbitrary side-shift nodes for boundary formulation. Using the proposed framework four different derivative matrix operators are introduced and their numerical stability is analyzed by studying their eigenvalues distribution in the complex plane. Their utility is studied by considering two important image processing problems, namely gradient surface reconstruction and image stitching. Experimentation indicates that the new derivative matrices not only outperform commonly used method but provide useful insights to the numerical issues in these two applications.
Recommendations
- Imaging and inverse problems of partial differential equations
- Inverse Problems Involving PDEs with Applications to Imaging
- scientific article; zbMATH DE number 750824
- Computational Methods for Inverse Problems in Imaging
- Discrete calculus, optimisation and inverse problems in imaging
- Inverse problems with inexact forward operator: iterative regularization and application in dynamic imaging
- Inverse problems and parameter identification in image processing
- scientific article; zbMATH DE number 800787
- Mimetic finite difference methods in image processing
- Discrete total variation with finite elements and applications to imaging
Cites work
- scientific article; zbMATH DE number 5599306 (Why is no real title available?)
- scientific article; zbMATH DE number 3179793 (Why is no real title available?)
- scientific article; zbMATH DE number 48170 (Why is no real title available?)
- scientific article; zbMATH DE number 2212009 (Why is no real title available?)
- A New Formula for Bivariate Hermite Interpolation on Variable Step Grids and Its Application to Image Interpolation
- A Review of Image Denoising Algorithms, with a New One
- A combined first and second order variational approach for image reconstruction
- A new look at partial fraction expansion from a high-level language viewpoint
- A single-lobe photometric stereo approach for heterogeneous material
- Algorithm 432 [C2]: Solution of the matrix equation AX + XB = C [F4]
- An algorithm for the finite difference approximation of derivatives with arbitrary degree and order of accuracy
- B-spline signal processing. I. Theory
- B-spline signal processing. II. Efficiency design and applications
- Chromatic Derivatives and Local Approximations
- Classroom Note:Calculation of Weights in Finite Difference Formulas
- Closed-form expressions for the finite difference approximations of first and higher derivatives based on Taylor series
- Computer Vision - ECCV 2004
- Direct analytical methods for solving Poisson equations in computer vision problems
- Discrete Lanczos derivatives of noisy data
- Finite difference weights, spectral differentiation, and superconvergence
- General explicit difference formulas for numerical differentiation
- Generation of Finite Difference Formulas on Arbitrarily Spaced Grids
- Generation of Pseudospectral Differentiation Matrices I
- Gradient-Based Image Recovery Methods From Incomplete Fourier Measurements
- Image alignment and stitching: a tutorial.
- Implicit image differentiation and filtering with applications to image sharpening
- Indoor Scene Reconstruction Using Near-Light Photometric Stereo
- Lanczo's Generalized Derivative
- Mathematical proof of closed form expressions for finite difference approximations based on Taylor series
- Matrix algorithms. Vol. 2: Eigensystems
- Multidimensional chromatic derivatives and series expansions
- New finite difference formulas for numerical differentiation
- Numerical differentiation with annihilators in noisy environment
- On the inversion of the Vandermonde matrix
- Regularized reconstruction of a surface from its measured gradient field
- Robust surface reconstruction
- Sharpening the response of a symmetric nonrecursive filter by multiple use of the same filter
- Style Transfer Via Texture Synthesis
- Taylor series based finite difference approximations of higher-degree derivatives
- The Least-Squares Property of the Lanczos Derivative
- The variational approach to shape from shading
- Total generalized variation
Cited in
(3)
This page was built for publication: Finite Differences in Forward and Inverse Imaging Problems: MaxPol Design
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4689649)
