Permuted Coordinate-Wise Optimizations Applied to Lp-Regularized Image Deconvolution

Jaeduk Han, Ki Sun Song, Jonghyun Kim, Moon Gi Kang

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Image deconvolution is an ill-posed problem that usually requires prior knowledge for regularizing the feasible solutions. In the literature, iterative methods estimate an intrinsic image, minimizing a cost function regularized by specific prior information. However, it is difficult to directly minimize the constrained cost function, if a nondifferentiable regularization (e.g., the sparsity constraint) is employed. In this paper, we propose a nonderivative image deconvolution algorithm that solves the under-constrained problem (i.e., a non-blind image deconvolution) by successively solving the permuted subproblems. The subproblems, arranged in permuted sequences, directly minimize the nondifferentiable cost functions. Various Lp-regularized (0<p ≤1, p=2) objective functions are utilized to demonstrate the pixel-wise optimization, in which the projection operator generates simplified, low-dimensional subproblems for estimating each pixel. The subproblems, after projection, are dealt within the corresponding hyperplanes containing the adjacent pixels of each image coordinate. Furthermore, successively solving the subproblems can accelerate the deconvolution process with a linear speedup, by parallelizing the subproblem sequences. The image deconvolution results with various regularization functionals are presented and the linear speedup is also demonstrated with a parallelized version of the proposed algorithm. Experimental results demonstrate that the proposed method outperforms the conventional methods in terms of the improved signal-to-noise ratio and structural similarity index measure.

Original languageEnglish
Pages (from-to)3556-3570
Number of pages15
JournalIEEE Transactions on Image Processing
Volume27
Issue number7
DOIs
Publication statusPublished - 2018 Jul

Fingerprint

Deconvolution
Cost functions
Pixels
Iterative methods
Mathematical operators
Signal to noise ratio

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Graphics and Computer-Aided Design

Cite this

@article{b4cfbe3a35c34538aadc2cfb314f656e,
title = "Permuted Coordinate-Wise Optimizations Applied to Lp-Regularized Image Deconvolution",
abstract = "Image deconvolution is an ill-posed problem that usually requires prior knowledge for regularizing the feasible solutions. In the literature, iterative methods estimate an intrinsic image, minimizing a cost function regularized by specific prior information. However, it is difficult to directly minimize the constrained cost function, if a nondifferentiable regularization (e.g., the sparsity constraint) is employed. In this paper, we propose a nonderivative image deconvolution algorithm that solves the under-constrained problem (i.e., a non-blind image deconvolution) by successively solving the permuted subproblems. The subproblems, arranged in permuted sequences, directly minimize the nondifferentiable cost functions. Various Lp-regularized (0",
author = "Jaeduk Han and Song, {Ki Sun} and Jonghyun Kim and Kang, {Moon Gi}",
year = "2018",
month = "7",
doi = "10.1109/TIP.2018.2825112",
language = "English",
volume = "27",
pages = "3556--3570",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "7",

}

Permuted Coordinate-Wise Optimizations Applied to Lp-Regularized Image Deconvolution. / Han, Jaeduk; Song, Ki Sun; Kim, Jonghyun; Kang, Moon Gi.

In: IEEE Transactions on Image Processing, Vol. 27, No. 7, 07.2018, p. 3556-3570.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Permuted Coordinate-Wise Optimizations Applied to Lp-Regularized Image Deconvolution

AU - Han, Jaeduk

AU - Song, Ki Sun

AU - Kim, Jonghyun

AU - Kang, Moon Gi

PY - 2018/7

Y1 - 2018/7

N2 - Image deconvolution is an ill-posed problem that usually requires prior knowledge for regularizing the feasible solutions. In the literature, iterative methods estimate an intrinsic image, minimizing a cost function regularized by specific prior information. However, it is difficult to directly minimize the constrained cost function, if a nondifferentiable regularization (e.g., the sparsity constraint) is employed. In this paper, we propose a nonderivative image deconvolution algorithm that solves the under-constrained problem (i.e., a non-blind image deconvolution) by successively solving the permuted subproblems. The subproblems, arranged in permuted sequences, directly minimize the nondifferentiable cost functions. Various Lp-regularized (0

AB - Image deconvolution is an ill-posed problem that usually requires prior knowledge for regularizing the feasible solutions. In the literature, iterative methods estimate an intrinsic image, minimizing a cost function regularized by specific prior information. However, it is difficult to directly minimize the constrained cost function, if a nondifferentiable regularization (e.g., the sparsity constraint) is employed. In this paper, we propose a nonderivative image deconvolution algorithm that solves the under-constrained problem (i.e., a non-blind image deconvolution) by successively solving the permuted subproblems. The subproblems, arranged in permuted sequences, directly minimize the nondifferentiable cost functions. Various Lp-regularized (0

UR - http://www.scopus.com/inward/record.url?scp=85045191591&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85045191591&partnerID=8YFLogxK

U2 - 10.1109/TIP.2018.2825112

DO - 10.1109/TIP.2018.2825112

M3 - Article

C2 - 29993832

AN - SCOPUS:85045191591

VL - 27

SP - 3556

EP - 3570

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 7

ER -