Fast Domain Decomposition for Global Image Smoothing

Youngjung Kim, Dongbo Min, Bumsub Ham, Kwanghoon Sohn

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Edge-preserving smoothing (EPS) can be formulated as minimizing an objective function that consists of data and regularization terms. At the price of high-computational cost, this global EPS approach is more robust and versatile than a local one that typically has a form of weighted averaging. In this paper, we introduce an efficient decomposition-based method for global EPS that minimizes the objective function of L 2 data and (possibly non-smooth and non-convex) regularization terms in linear time. Different from previous decomposition-based methods, which require solving a large linear system, our approach solves an equivalent constrained optimization problem, resulting in a sequence of 1-D sub-problems. This enables applying fast linear time solver for weighted-least squares and - L 1 smoothing problems. An alternating direction method of multipliers algorithm is adopted to guarantee fast convergence. Our method is fully parallelizable, and its runtime is even comparable to the state-of-the-art local EPS approaches. We also propose a family of fast majorization-minimization algorithms that minimize an objective with non-convex regularization terms. Experimental results demonstrate the effectiveness and flexibility of our approach in a range of image processing and computational photography applications.

Original languageEnglish
Article number7937834
Pages (from-to)4079-4091
Number of pages13
JournalIEEE Transactions on Image Processing
Volume26
Issue number8
DOIs
Publication statusPublished - 2017 Aug 1

Fingerprint

Decomposition
Constrained optimization
Photography
Linear systems
Image processing
Costs

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Graphics and Computer-Aided Design

Cite this

@article{aed8d1026ab54ed18d6acf05d30555ef,
title = "Fast Domain Decomposition for Global Image Smoothing",
abstract = "Edge-preserving smoothing (EPS) can be formulated as minimizing an objective function that consists of data and regularization terms. At the price of high-computational cost, this global EPS approach is more robust and versatile than a local one that typically has a form of weighted averaging. In this paper, we introduce an efficient decomposition-based method for global EPS that minimizes the objective function of L 2 data and (possibly non-smooth and non-convex) regularization terms in linear time. Different from previous decomposition-based methods, which require solving a large linear system, our approach solves an equivalent constrained optimization problem, resulting in a sequence of 1-D sub-problems. This enables applying fast linear time solver for weighted-least squares and - L 1 smoothing problems. An alternating direction method of multipliers algorithm is adopted to guarantee fast convergence. Our method is fully parallelizable, and its runtime is even comparable to the state-of-the-art local EPS approaches. We also propose a family of fast majorization-minimization algorithms that minimize an objective with non-convex regularization terms. Experimental results demonstrate the effectiveness and flexibility of our approach in a range of image processing and computational photography applications.",
author = "Youngjung Kim and Dongbo Min and Bumsub Ham and Kwanghoon Sohn",
year = "2017",
month = "8",
day = "1",
doi = "10.1109/TIP.2017.2710621",
language = "English",
volume = "26",
pages = "4079--4091",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "8",

}

Fast Domain Decomposition for Global Image Smoothing. / Kim, Youngjung; Min, Dongbo; Ham, Bumsub; Sohn, Kwanghoon.

In: IEEE Transactions on Image Processing, Vol. 26, No. 8, 7937834, 01.08.2017, p. 4079-4091.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fast Domain Decomposition for Global Image Smoothing

AU - Kim, Youngjung

AU - Min, Dongbo

AU - Ham, Bumsub

AU - Sohn, Kwanghoon

PY - 2017/8/1

Y1 - 2017/8/1

N2 - Edge-preserving smoothing (EPS) can be formulated as minimizing an objective function that consists of data and regularization terms. At the price of high-computational cost, this global EPS approach is more robust and versatile than a local one that typically has a form of weighted averaging. In this paper, we introduce an efficient decomposition-based method for global EPS that minimizes the objective function of L 2 data and (possibly non-smooth and non-convex) regularization terms in linear time. Different from previous decomposition-based methods, which require solving a large linear system, our approach solves an equivalent constrained optimization problem, resulting in a sequence of 1-D sub-problems. This enables applying fast linear time solver for weighted-least squares and - L 1 smoothing problems. An alternating direction method of multipliers algorithm is adopted to guarantee fast convergence. Our method is fully parallelizable, and its runtime is even comparable to the state-of-the-art local EPS approaches. We also propose a family of fast majorization-minimization algorithms that minimize an objective with non-convex regularization terms. Experimental results demonstrate the effectiveness and flexibility of our approach in a range of image processing and computational photography applications.

AB - Edge-preserving smoothing (EPS) can be formulated as minimizing an objective function that consists of data and regularization terms. At the price of high-computational cost, this global EPS approach is more robust and versatile than a local one that typically has a form of weighted averaging. In this paper, we introduce an efficient decomposition-based method for global EPS that minimizes the objective function of L 2 data and (possibly non-smooth and non-convex) regularization terms in linear time. Different from previous decomposition-based methods, which require solving a large linear system, our approach solves an equivalent constrained optimization problem, resulting in a sequence of 1-D sub-problems. This enables applying fast linear time solver for weighted-least squares and - L 1 smoothing problems. An alternating direction method of multipliers algorithm is adopted to guarantee fast convergence. Our method is fully parallelizable, and its runtime is even comparable to the state-of-the-art local EPS approaches. We also propose a family of fast majorization-minimization algorithms that minimize an objective with non-convex regularization terms. Experimental results demonstrate the effectiveness and flexibility of our approach in a range of image processing and computational photography applications.

UR - http://www.scopus.com/inward/record.url?scp=85028422835&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028422835&partnerID=8YFLogxK

U2 - 10.1109/TIP.2017.2710621

DO - 10.1109/TIP.2017.2710621

M3 - Article

VL - 26

SP - 4079

EP - 4091

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 8

M1 - 7937834

ER -