Global active contour-based image segmentation via probability alignment

Andriy Myronenko, Xubo Song

Research output: Chapter in Book/Report/Conference proceedingConference contribution

16 Citations (Scopus)

Abstract

Active contours is a popular technique for image segmentation. However, active contour tend to converge to the closest local minimum of its energy function and often requires a close boundary initialization. We introduce a new approach that overcomes the close boundary initialization problem by reformulating the external energy term. We treat the active contour as a mean curve of the probability density function p(x). It moves to minimize the Kullback-Leibler (KL) divergence between p(x) and the probability density function derived from the image. KL divergence forces p(x) to .cover all image areas. and the uncovered areas are heavily penalized, which allows the active contour to go over the edges. Also we use deterministic annealing on the width of p(x) to implement a coarse-to-fine search strategy. In the limit, when the width of p(x) goes to zero, the KL divergence function converges to the conventional external energy term (which can be seen a special case) of active contours. Our method produces robust segmentation results from arbitrary initialization positions.

Original languageEnglish (US)
Title of host publication2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009
Pages2798-2804
Number of pages7
DOIs
StatePublished - 2009
Event2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009 - Miami, FL, United States
Duration: Jun 20 2009Jun 25 2009

Other

Other2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009
CountryUnited States
CityMiami, FL
Period6/20/096/25/09

Fingerprint

Image segmentation
Probability density function
Annealing

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Biomedical Engineering

Cite this

Myronenko, A., & Song, X. (2009). Global active contour-based image segmentation via probability alignment. In 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009 (pp. 2798-2804). [5206552] https://doi.org/10.1109/CVPRW.2009.5206552

Global active contour-based image segmentation via probability alignment. / Myronenko, Andriy; Song, Xubo.

2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009. 2009. p. 2798-2804 5206552.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Myronenko, A & Song, X 2009, Global active contour-based image segmentation via probability alignment. in 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009., 5206552, pp. 2798-2804, 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009, Miami, FL, United States, 6/20/09. https://doi.org/10.1109/CVPRW.2009.5206552
Myronenko A, Song X. Global active contour-based image segmentation via probability alignment. In 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009. 2009. p. 2798-2804. 5206552 https://doi.org/10.1109/CVPRW.2009.5206552
Myronenko, Andriy ; Song, Xubo. / Global active contour-based image segmentation via probability alignment. 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009. 2009. pp. 2798-2804
@inproceedings{16bfdb93dcca443f9d57f46f4bf59678,
title = "Global active contour-based image segmentation via probability alignment",
abstract = "Active contours is a popular technique for image segmentation. However, active contour tend to converge to the closest local minimum of its energy function and often requires a close boundary initialization. We introduce a new approach that overcomes the close boundary initialization problem by reformulating the external energy term. We treat the active contour as a mean curve of the probability density function p(x). It moves to minimize the Kullback-Leibler (KL) divergence between p(x) and the probability density function derived from the image. KL divergence forces p(x) to .cover all image areas. and the uncovered areas are heavily penalized, which allows the active contour to go over the edges. Also we use deterministic annealing on the width of p(x) to implement a coarse-to-fine search strategy. In the limit, when the width of p(x) goes to zero, the KL divergence function converges to the conventional external energy term (which can be seen a special case) of active contours. Our method produces robust segmentation results from arbitrary initialization positions.",
author = "Andriy Myronenko and Xubo Song",
year = "2009",
doi = "10.1109/CVPRW.2009.5206552",
language = "English (US)",
isbn = "9781424439935",
pages = "2798--2804",
booktitle = "2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009",

}

TY - GEN

T1 - Global active contour-based image segmentation via probability alignment

AU - Myronenko, Andriy

AU - Song, Xubo

PY - 2009

Y1 - 2009

N2 - Active contours is a popular technique for image segmentation. However, active contour tend to converge to the closest local minimum of its energy function and often requires a close boundary initialization. We introduce a new approach that overcomes the close boundary initialization problem by reformulating the external energy term. We treat the active contour as a mean curve of the probability density function p(x). It moves to minimize the Kullback-Leibler (KL) divergence between p(x) and the probability density function derived from the image. KL divergence forces p(x) to .cover all image areas. and the uncovered areas are heavily penalized, which allows the active contour to go over the edges. Also we use deterministic annealing on the width of p(x) to implement a coarse-to-fine search strategy. In the limit, when the width of p(x) goes to zero, the KL divergence function converges to the conventional external energy term (which can be seen a special case) of active contours. Our method produces robust segmentation results from arbitrary initialization positions.

AB - Active contours is a popular technique for image segmentation. However, active contour tend to converge to the closest local minimum of its energy function and often requires a close boundary initialization. We introduce a new approach that overcomes the close boundary initialization problem by reformulating the external energy term. We treat the active contour as a mean curve of the probability density function p(x). It moves to minimize the Kullback-Leibler (KL) divergence between p(x) and the probability density function derived from the image. KL divergence forces p(x) to .cover all image areas. and the uncovered areas are heavily penalized, which allows the active contour to go over the edges. Also we use deterministic annealing on the width of p(x) to implement a coarse-to-fine search strategy. In the limit, when the width of p(x) goes to zero, the KL divergence function converges to the conventional external energy term (which can be seen a special case) of active contours. Our method produces robust segmentation results from arbitrary initialization positions.

UR - http://www.scopus.com/inward/record.url?scp=70450159370&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70450159370&partnerID=8YFLogxK

U2 - 10.1109/CVPRW.2009.5206552

DO - 10.1109/CVPRW.2009.5206552

M3 - Conference contribution

AN - SCOPUS:70450159370

SN - 9781424439935

SP - 2798

EP - 2804

BT - 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009

ER -