Tracking non-rigid objects using probabilistic Hausdorff distance matching

Sang Cheol Park, Sung Hoon Lim, Bong K. Sin, Seong Whan Lee

Research output: Contribution to journalArticle

27 Citations (Scopus)

Abstract

This paper proposes a new method of extracting and tracking a non-rigid object moving against a cluttered background while allowing camera movement. For object extraction we first detect an object using watershed segmentation technique and then extract its contour points by approximating the boundary using the idea of feature point weighting. For object tracking we take the contour to estimate its motion in the next frame by the maximum likelihood method. The position of the object is estimated using a probabilistic Hausdorff measurement while the shape variation is modelled using a modified active contour model. The proposed method is highly tolerant to occlusion. Unless an object is fully occluded during tracking, the result is stable and the method is robust enough for practical application.

Original languageEnglish
Pages (from-to)2373-2384
Number of pages12
JournalPattern Recognition
Volume38
Issue number12
DOIs
Publication statusPublished - 2005 Dec 1

Fingerprint

Watersheds
Maximum likelihood
Cameras

Keywords

  • Active contour
  • Hausdorff distance
  • Object tracking
  • Watershed segmentation

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

Tracking non-rigid objects using probabilistic Hausdorff distance matching. / Park, Sang Cheol; Lim, Sung Hoon; Sin, Bong K.; Lee, Seong Whan.

In: Pattern Recognition, Vol. 38, No. 12, 01.12.2005, p. 2373-2384.

Research output: Contribution to journalArticle

Park, Sang Cheol ; Lim, Sung Hoon ; Sin, Bong K. ; Lee, Seong Whan. / Tracking non-rigid objects using probabilistic Hausdorff distance matching. In: Pattern Recognition. 2005 ; Vol. 38, No. 12. pp. 2373-2384.
@article{c17506bfc75b4ad49e07d5fcfc7aaabf,
title = "Tracking non-rigid objects using probabilistic Hausdorff distance matching",
abstract = "This paper proposes a new method of extracting and tracking a non-rigid object moving against a cluttered background while allowing camera movement. For object extraction we first detect an object using watershed segmentation technique and then extract its contour points by approximating the boundary using the idea of feature point weighting. For object tracking we take the contour to estimate its motion in the next frame by the maximum likelihood method. The position of the object is estimated using a probabilistic Hausdorff measurement while the shape variation is modelled using a modified active contour model. The proposed method is highly tolerant to occlusion. Unless an object is fully occluded during tracking, the result is stable and the method is robust enough for practical application.",
keywords = "Active contour, Hausdorff distance, Object tracking, Watershed segmentation",
author = "Park, {Sang Cheol} and Lim, {Sung Hoon} and Sin, {Bong K.} and Lee, {Seong Whan}",
year = "2005",
month = "12",
day = "1",
doi = "10.1016/j.patcog.2005.01.015",
language = "English",
volume = "38",
pages = "2373--2384",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",
number = "12",

}

TY - JOUR

T1 - Tracking non-rigid objects using probabilistic Hausdorff distance matching

AU - Park, Sang Cheol

AU - Lim, Sung Hoon

AU - Sin, Bong K.

AU - Lee, Seong Whan

PY - 2005/12/1

Y1 - 2005/12/1

N2 - This paper proposes a new method of extracting and tracking a non-rigid object moving against a cluttered background while allowing camera movement. For object extraction we first detect an object using watershed segmentation technique and then extract its contour points by approximating the boundary using the idea of feature point weighting. For object tracking we take the contour to estimate its motion in the next frame by the maximum likelihood method. The position of the object is estimated using a probabilistic Hausdorff measurement while the shape variation is modelled using a modified active contour model. The proposed method is highly tolerant to occlusion. Unless an object is fully occluded during tracking, the result is stable and the method is robust enough for practical application.

AB - This paper proposes a new method of extracting and tracking a non-rigid object moving against a cluttered background while allowing camera movement. For object extraction we first detect an object using watershed segmentation technique and then extract its contour points by approximating the boundary using the idea of feature point weighting. For object tracking we take the contour to estimate its motion in the next frame by the maximum likelihood method. The position of the object is estimated using a probabilistic Hausdorff measurement while the shape variation is modelled using a modified active contour model. The proposed method is highly tolerant to occlusion. Unless an object is fully occluded during tracking, the result is stable and the method is robust enough for practical application.

KW - Active contour

KW - Hausdorff distance

KW - Object tracking

KW - Watershed segmentation

UR - http://www.scopus.com/inward/record.url?scp=25144482489&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=25144482489&partnerID=8YFLogxK

U2 - 10.1016/j.patcog.2005.01.015

DO - 10.1016/j.patcog.2005.01.015

M3 - Article

AN - SCOPUS:25144482489

VL - 38

SP - 2373

EP - 2384

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 12

ER -