Image completion using planar structure guidance

Jia Bin Huang, Sing Bing Kang, Narendra Ahuja, Johannes Kopf

Research output: Contribution to journalConference article

Abstract

We propose a method for automatically guiding patch-based image completion using mid-level structural cues. Our method first estimates planar projection parameters, softly segments the known region into planes, and discovers translational regularity within these planes. This information is then converted into soft constraints for the low-level completion algorithm by defining prior probabilities for patch offsets and transformations. Our method handles multiple planes, and in the absence of any detected planes falls back to a baseline fronto-parallel image completion algorithm. We validate our technique through extensive comparisons with state-of-the-art algorithms on a variety of scenes.

Original languageEnglish (US)
Article number129
JournalACM Transactions on Graphics
Volume33
Issue number4
DOIs
StatePublished - Jan 1 2014
Event41st International Conference and Exhibition on Computer Graphics and Interactive Techniques, ACM SIGGRAPH 2014 - Vancouver, BC, Canada
Duration: Aug 10 2014Aug 14 2014

Keywords

  • Guided synthesis
  • Image completion
  • Mid-level analysis
  • Patch-based synthesis

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design

Cite this

Image completion using planar structure guidance. / Huang, Jia Bin; Kang, Sing Bing; Ahuja, Narendra; Kopf, Johannes.

In: ACM Transactions on Graphics, Vol. 33, No. 4, 129, 01.01.2014.

Research output: Contribution to journalConference article

Huang, Jia Bin ; Kang, Sing Bing ; Ahuja, Narendra ; Kopf, Johannes. / Image completion using planar structure guidance. In: ACM Transactions on Graphics. 2014 ; Vol. 33, No. 4.
@article{31684182f1ed4f208149d89b6c651579,
title = "Image completion using planar structure guidance",
abstract = "We propose a method for automatically guiding patch-based image completion using mid-level structural cues. Our method first estimates planar projection parameters, softly segments the known region into planes, and discovers translational regularity within these planes. This information is then converted into soft constraints for the low-level completion algorithm by defining prior probabilities for patch offsets and transformations. Our method handles multiple planes, and in the absence of any detected planes falls back to a baseline fronto-parallel image completion algorithm. We validate our technique through extensive comparisons with state-of-the-art algorithms on a variety of scenes.",
keywords = "Guided synthesis, Image completion, Mid-level analysis, Patch-based synthesis",
author = "Huang, {Jia Bin} and Kang, {Sing Bing} and Narendra Ahuja and Johannes Kopf",
year = "2014",
month = "1",
day = "1",
doi = "10.1145/2601097.2601205",
language = "English (US)",
volume = "33",
journal = "ACM Transactions on Computer Systems",
issn = "0730-0301",
publisher = "Association for Computing Machinery (ACM)",
number = "4",

}

TY - JOUR

T1 - Image completion using planar structure guidance

AU - Huang, Jia Bin

AU - Kang, Sing Bing

AU - Ahuja, Narendra

AU - Kopf, Johannes

PY - 2014/1/1

Y1 - 2014/1/1

N2 - We propose a method for automatically guiding patch-based image completion using mid-level structural cues. Our method first estimates planar projection parameters, softly segments the known region into planes, and discovers translational regularity within these planes. This information is then converted into soft constraints for the low-level completion algorithm by defining prior probabilities for patch offsets and transformations. Our method handles multiple planes, and in the absence of any detected planes falls back to a baseline fronto-parallel image completion algorithm. We validate our technique through extensive comparisons with state-of-the-art algorithms on a variety of scenes.

AB - We propose a method for automatically guiding patch-based image completion using mid-level structural cues. Our method first estimates planar projection parameters, softly segments the known region into planes, and discovers translational regularity within these planes. This information is then converted into soft constraints for the low-level completion algorithm by defining prior probabilities for patch offsets and transformations. Our method handles multiple planes, and in the absence of any detected planes falls back to a baseline fronto-parallel image completion algorithm. We validate our technique through extensive comparisons with state-of-the-art algorithms on a variety of scenes.

KW - Guided synthesis

KW - Image completion

KW - Mid-level analysis

KW - Patch-based synthesis

UR - http://www.scopus.com/inward/record.url?scp=84905743116&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84905743116&partnerID=8YFLogxK

U2 - 10.1145/2601097.2601205

DO - 10.1145/2601097.2601205

M3 - Conference article

AN - SCOPUS:84905743116

VL - 33

JO - ACM Transactions on Computer Systems

JF - ACM Transactions on Computer Systems

SN - 0730-0301

IS - 4

M1 - 129

ER -