Spectral Methods for Passive Imaging: Nonasymptotic Performance and Robustness
From MaRDI portal
Abstract: We study the problem of passive imaging through convolutive channels. A scene is illuminated with an unknown, unstructured source, and the measured response is the convolution of this source with multiple channel responses, each of which is time-limited. Spectral methods based on the commutativity of convolution, first proposed and analyzed in the 1990s, provide an elegant mathematical framework for attacking this problem. However, these now classical methods are very sensitive to noise, especially when working from relatively small sample sizes. In this paper, we show that a linear subspace model on the coefficients of the impulse responses of the channels can make this problem well-posed. We derive non-asymptotic error bounds for the generic subspace model by analyzing the spectral gap of the cross-correlation matrix of the channels relative to the perturbation introduced by noise. Numerical results show that this modified spectral method offers significant improvements over the classical method and outperforms other competing methods for multichannel blind deconvolution.
Recommendations
- A Parametric Estimation Approach to Instantaneous Spectral Imaging
- Passive Sensor Imaging Using Cross Correlations of Noisy Signals in a Scattering Medium
- Passive imaging using distributed apertures in multiple-scattering environments
- Passive synthetic aperture imaging with limited noise sources
- Computational spectral and ultrafast imaging via convex optimization
- Spectral imaging based on 2D diffraction patterns and robust principal component analysis
- Signal-to-noise ratio estimation in passive correlation-based imaging
- Sparse Signal Recovery Based Imaging in the Presence of Mode Conversion With Application to Non-Destructive Testing
Cites work
- A mathematical introduction to compressive sensing
- Adaptive estimation of a quadratic functional by model selection.
- Blind Deconvolution Using Convex Programming
- Blind identification of multichannel FIR blurs and perfect image restoration
- Blind Recovery of Sparse Signals From Subsampled Convolution
- Block diagonally dominant matrices and generalizations of the Gerschgorin circle theorem
- Duality of metric entropy
- Hanson-Wright inequality and sub-Gaussian concentration
- scientific article; zbMATH DE number 6159604 (Why is no real title available?)
- Identifiability and Stability in Blind Deconvolution Under Minimal Assumptions
- Identifiability in Blind Deconvolution With Subspace or Sparsity Constraints
- Inequalities of Bernstein-Jackson-type and the degree of compactness of operators in Banach spaces
- Local operator theory, random matrices and Banach spaces.
- MCA: A Multichannel Approach to SAR Autofocus
- Noncommutative Bennett and Rosenthal inequalities
- Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
- Optimal injectivity conditions for bilinear inverse problems with applications to identifiability of deconvolution problems
- Optimal Sample Complexity for Blind Gain and Phase Calibration
- Passive synthetic aperture imaging
- Random perturbation of low rank matrices: improving classical bounds
- Rapid, robust, and reliable blind deconvolution via nonconvex optimization
- Recent developments in blind channel equalization: From cyclostationarity to subspaces
- Self-calibration and bilinear inverse problems via linear least squares
- Singular vectors under random perturbation
- Suprema of chaos processes and the restricted isometry property
- The Rotation of Eigenvectors by a Perturbation. III
- User-friendly tail bounds for sums of random matrices
Cited in
(4)- Sparse power factorization: balancing peakiness and sample complexity
- Passive imaging using distributed apertures in multiple-scattering environments
- Exact Recovery of Multichannel Sparse Blind Deconvolution via Gradient Descent
- Non-asymptotic behavior of the spectrum of the sinc-kernel operator and related applications
This page was built for publication: Spectral Methods for Passive Imaging: Nonasymptotic Performance and Robustness
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5230405)
