Field of Particle Filters Image Inpainting

Anne Cuzol; Kim Steenstrup Pedersen; Mads Nielsen

doi:10.1007/s10851-008-0072-7

Field of Particle Filters Image Inpainting

Anne Cuzol, Kim Steenstrup Pedersen, Mads Nielsen

3 Citations (Scopus)

Abstract

We present a novel algorithm for solving the image
inpainting problem based on a field of locally interacting
particle filters. Image inpainting, also known as image
completion, is concerned with the problem of filling image
regions with new visually plausible data. In order to avoid
the difficulty of solving the problem globally for the region
to be inpainted, we introduce a field of local particle
filters. The states of the particle filters are image patches.
Global consistency is enforced by a Markov random field
image model which connects neighbouring particle filters.
The benefit of using locally interacting particle filters is that
several competing hypotheses on inpainting solutions are
kept active, allowing the method to provide globally consistent
solutions on problems where other local methods may
fail. We provide examples of applications of the developed
method.

Keywords: Inpainting · Image completion · Hole filling ·
Particle filter · Markov random field

Original language	English
Journal	Journal of Mathematical Imaging and Vision
Volume	31
Issue number	2-3
Pages (from-to)	147-156
Number of pages	10
ISSN	0924-9907
DOIs	https://doi.org/10.1007/s10851-008-0072-7
Publication status	Published - 2008

Keywords

Faculty of Science
Inpainting
Image completion
Hole filling
Particle filter
Markov random field

Access to Document

10.1007/s10851-008-0072-7

Cite this

@article{da9c0030a01f11dd86a6000ea68e967b,

title = "Field of Particle Filters Image Inpainting",

abstract = "We present a novel algorithm for solving the imageinpainting problem based on a field of locally interactingparticle filters. Image inpainting, also known as imagecompletion, is concerned with the problem of filling imageregions with new visually plausible data. In order to avoidthe difficulty of solving the problem globally for the regionto be inpainted, we introduce a field of local particlefilters. The states of the particle filters are image patches.Global consistency is enforced by a Markov random fieldimage model which connects neighbouring particle filters.The benefit of using locally interacting particle filters is thatseveral competing hypotheses on inpainting solutions arekept active, allowing the method to provide globally consistentsolutions on problems where other local methods mayfail. We provide examples of applications of the developedmethod.Keywords: Inpainting · Image completion · Hole filling ·Particle filter · Markov random field",

keywords = "Faculty of Science, Inpainting, Image completion, Hole filling, Particle filter, Markov random field",

author = "Anne Cuzol and Pedersen, {Kim Steenstrup} and Mads Nielsen",

year = "2008",

doi = "10.1007/s10851-008-0072-7",

language = "English",

volume = "31",

pages = "147--156",

journal = "Journal of Mathematical Imaging and Vision",

issn = "0924-9907",

publisher = "Springer",

number = "2-3",

}

TY - JOUR

T1 - Field of Particle Filters Image Inpainting

AU - Cuzol, Anne

AU - Pedersen, Kim Steenstrup

AU - Nielsen, Mads

PY - 2008

Y1 - 2008

N2 - We present a novel algorithm for solving the imageinpainting problem based on a field of locally interactingparticle filters. Image inpainting, also known as imagecompletion, is concerned with the problem of filling imageregions with new visually plausible data. In order to avoidthe difficulty of solving the problem globally for the regionto be inpainted, we introduce a field of local particlefilters. The states of the particle filters are image patches.Global consistency is enforced by a Markov random fieldimage model which connects neighbouring particle filters.The benefit of using locally interacting particle filters is thatseveral competing hypotheses on inpainting solutions arekept active, allowing the method to provide globally consistentsolutions on problems where other local methods mayfail. We provide examples of applications of the developedmethod.Keywords: Inpainting · Image completion · Hole filling ·Particle filter · Markov random field

AB - We present a novel algorithm for solving the imageinpainting problem based on a field of locally interactingparticle filters. Image inpainting, also known as imagecompletion, is concerned with the problem of filling imageregions with new visually plausible data. In order to avoidthe difficulty of solving the problem globally for the regionto be inpainted, we introduce a field of local particlefilters. The states of the particle filters are image patches.Global consistency is enforced by a Markov random fieldimage model which connects neighbouring particle filters.The benefit of using locally interacting particle filters is thatseveral competing hypotheses on inpainting solutions arekept active, allowing the method to provide globally consistentsolutions on problems where other local methods mayfail. We provide examples of applications of the developedmethod.Keywords: Inpainting · Image completion · Hole filling ·Particle filter · Markov random field

KW - Faculty of Science

KW - Inpainting

KW - Image completion

KW - Hole filling

KW - Particle filter

KW - Markov random field

U2 - 10.1007/s10851-008-0072-7

DO - 10.1007/s10851-008-0072-7

M3 - Journal article

SN - 0924-9907

VL - 31

SP - 147

EP - 156

JO - Journal of Mathematical Imaging and Vision

JF - Journal of Mathematical Imaging and Vision

IS - 2-3

ER -

Field of Particle Filters Image Inpainting

Abstract

Keywords

Access to Document

Fingerprint

Cite this