File Download
 
Links for fulltext
(May Require Subscription)
 
Supplementary

Article: Interactive image segmentation based on level sets of probabilities
  • Basic View
  • Metadata View
  • XML View
TitleInteractive image segmentation based on level sets of probabilities
 
AuthorsLiu, Y1
Yu, Y2
 
KeywordsCurvature
Distance Transform
Image Segmentation
Level Set Method
Statistical Classification
 
Issue Date2012
 
PublisherI E E E. The Journal's web site is located at http://www.computer.org/tvcg
 
CitationIeee Transactions On Visualization And Computer Graphics, 2012, v. 18 n. 2, p. 202-213 [How to Cite?]
DOI: http://dx.doi.org/10.1109/TVCG.2011.77
 
AbstractIn this paper, we present a robust and accurate algorithm for interactive image segmentation. The level set method is clearly advantageous for image objects with a complex topology and fragmented appearance. Our method integrates discriminative classification models and distance transforms with the level set method to avoid local minima and better snap to true object boundaries. The level set function approximates a transformed version of pixelwise posterior probabilities of being part of a target object. The evolution of its zero level set is driven by three force terms, region force, edge field force, and curvature force. These forces are based on a probabilistic classifier and an unsigned distance transform of salient edges. We further propose a technique that improves the performance of both the probabilistic classifier and the level set method over multiple passes. It makes the final object segmentation less sensitive to user interactions. Experiments and comparisons demonstrate the effectiveness of our method. © 2012 IEEE.
 
ISSN1077-2626
2012 Impact Factor: 1.898
2012 SCImago Journal Rankings: 1.890
 
DOIhttp://dx.doi.org/10.1109/TVCG.2011.77
 
ISI Accession Number IDWOS:000298043100003
Funding AgencyGrant Number
US National Science Foundation (NSF)IIS 09-14631
Funding Information:

We would like to thank the reviewers, whose comments have been valuable in improving our manuscript. This work was partially supported by US National Science Foundation (NSF) (IIS 09-14631).

 
ReferencesReferences in Scopus
 
DC FieldValue
dc.contributor.authorLiu, Y
 
dc.contributor.authorYu, Y
 
dc.date.accessioned2012-06-26T06:39:35Z
 
dc.date.available2012-06-26T06:39:35Z
 
dc.date.issued2012
 
dc.description.abstractIn this paper, we present a robust and accurate algorithm for interactive image segmentation. The level set method is clearly advantageous for image objects with a complex topology and fragmented appearance. Our method integrates discriminative classification models and distance transforms with the level set method to avoid local minima and better snap to true object boundaries. The level set function approximates a transformed version of pixelwise posterior probabilities of being part of a target object. The evolution of its zero level set is driven by three force terms, region force, edge field force, and curvature force. These forces are based on a probabilistic classifier and an unsigned distance transform of salient edges. We further propose a technique that improves the performance of both the probabilistic classifier and the level set method over multiple passes. It makes the final object segmentation less sensitive to user interactions. Experiments and comparisons demonstrate the effectiveness of our method. © 2012 IEEE.
 
dc.description.naturepublished_or_final_version
 
dc.identifier.citationIeee Transactions On Visualization And Computer Graphics, 2012, v. 18 n. 2, p. 202-213 [How to Cite?]
DOI: http://dx.doi.org/10.1109/TVCG.2011.77
 
dc.identifier.doihttp://dx.doi.org/10.1109/TVCG.2011.77
 
dc.identifier.epage213
 
dc.identifier.hkuros200757
 
dc.identifier.isiWOS:000298043100003
Funding AgencyGrant Number
US National Science Foundation (NSF)IIS 09-14631
Funding Information:

We would like to thank the reviewers, whose comments have been valuable in improving our manuscript. This work was partially supported by US National Science Foundation (NSF) (IIS 09-14631).

 
dc.identifier.issn1077-2626
2012 Impact Factor: 1.898
2012 SCImago Journal Rankings: 1.890
 
dc.identifier.issue2
 
dc.identifier.scopuseid_2-s2.0-83855162158
 
dc.identifier.spage202
 
dc.identifier.urihttp://hdl.handle.net/10722/152486
 
dc.identifier.volume18
 
dc.languageeng
 
dc.publisherI E E E. The Journal's web site is located at http://www.computer.org/tvcg
 
dc.publisher.placeUnited States
 
dc.relation.ispartofIEEE Transactions on Visualization and Computer Graphics
 
dc.relation.referencesReferences in Scopus
 
dc.rightsCreative Commons: Attribution 3.0 Hong Kong License
 
dc.subjectCurvature
 
dc.subjectDistance Transform
 
dc.subjectImage Segmentation
 
dc.subjectLevel Set Method
 
dc.subjectStatistical Classification
 
dc.titleInteractive image segmentation based on level sets of probabilities
 
dc.typeArticle
 
<?xml encoding="utf-8" version="1.0"?>
<item><contributor.author>Liu, Y</contributor.author>
<contributor.author>Yu, Y</contributor.author>
<date.accessioned>2012-06-26T06:39:35Z</date.accessioned>
<date.available>2012-06-26T06:39:35Z</date.available>
<date.issued>2012</date.issued>
<identifier.citation>Ieee Transactions On Visualization And Computer Graphics, 2012, v. 18 n. 2, p. 202-213</identifier.citation>
<identifier.issn>1077-2626</identifier.issn>
<identifier.uri>http://hdl.handle.net/10722/152486</identifier.uri>
<description.abstract>In this paper, we present a robust and accurate algorithm for interactive image segmentation. The level set method is clearly advantageous for image objects with a complex topology and fragmented appearance. Our method integrates discriminative classification models and distance transforms with the level set method to avoid local minima and better snap to true object boundaries. The level set function approximates a transformed version of pixelwise posterior probabilities of being part of a target object. The evolution of its zero level set is driven by three force terms, region force, edge field force, and curvature force. These forces are based on a probabilistic classifier and an unsigned distance transform of salient edges. We further propose a technique that improves the performance of both the probabilistic classifier and the level set method over multiple passes. It makes the final object segmentation less sensitive to user interactions. Experiments and comparisons demonstrate the effectiveness of our method. &#169; 2012 IEEE.</description.abstract>
<language>eng</language>
<publisher>I E E E. The Journal&apos;s web site is located at http://www.computer.org/tvcg</publisher>
<relation.ispartof>IEEE Transactions on Visualization and Computer Graphics</relation.ispartof>
<rights>Creative Commons: Attribution 3.0 Hong Kong License</rights>
<subject>Curvature</subject>
<subject>Distance Transform</subject>
<subject>Image Segmentation</subject>
<subject>Level Set Method</subject>
<subject>Statistical Classification</subject>
<title>Interactive image segmentation based on level sets of probabilities</title>
<type>Article</type>
<description.nature>published_or_final_version</description.nature>
<identifier.doi>10.1109/TVCG.2011.77</identifier.doi>
<identifier.scopus>eid_2-s2.0-83855162158</identifier.scopus>
<identifier.hkuros>200757</identifier.hkuros>
<relation.references>http://www.scopus.com/mlt/select.url?eid=2-s2.0-83855162158&amp;selection=ref&amp;src=s&amp;origin=recordpage</relation.references>
<identifier.volume>18</identifier.volume>
<identifier.issue>2</identifier.issue>
<identifier.spage>202</identifier.spage>
<identifier.epage>213</identifier.epage>
<identifier.isi>WOS:000298043100003</identifier.isi>
<publisher.place>United States</publisher.place>
<bitstream.url>http://hub.hku.hk/bitstream/10722/152486/1/Content.pdf</bitstream.url>
</item>
Author Affiliations
  1. University of Electronic Science and Technology of China
  2. University of Illinois at Urbana-Champaign