Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion

Junsoo Kim, Yoonsik Choe, Yong Goo Kim

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

3 Citations (Scopus)

Abstract

Generation of object-boundary-matched accurate depth map is important to semi-automatic 2D to 3D conversion for high quality perception of converted 3D video. In this paper, we propose an accurate depth map generation scheme using an MRF-based initial contour tracking followed by a robust graph-cut-based contour refinement. To limit the propagation of potential tracking errors, key-frames as well as current and previous video frames are considered together in composing the graph cut seeds of the proposed contour refinement. In addition, the proposed scheme checks the ambiguity of seeds in foreground and background to eliminate the seeds, which may cause confusion in graph-cut segmentation, and thereby further improves the accuracy of contour tracking. Experiment results demonstrate the proposed scheme is especially useful when there is complex background or deformation of object, indicating the power of the proposed scheme in object-boundary-matched accurate depth map generation.

Original languageEnglish
Title of host publication2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011
Pages358-363
Number of pages6
DOIs
Publication statusPublished - 2011 Oct 24
Event13th IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011 - Victoria, BC, Canada
Duration: 2011 Aug 232011 Aug 26

Publication series

NameIEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings

Other

Other13th IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011
CountryCanada
CityVictoria, BC
Period11/8/2311/8/26

Fingerprint

Seed
Experiments

All Science Journal Classification (ASJC) codes

  • Signal Processing
  • Computer Networks and Communications

Cite this

Kim, J., Choe, Y., & Kim, Y. G. (2011). Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion. In 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011 (pp. 358-363). [6032919] (IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings). https://doi.org/10.1109/PACRIM.2011.6032919
Kim, Junsoo ; Choe, Yoonsik ; Kim, Yong Goo. / Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion. 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011. 2011. pp. 358-363 (IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings).
@inproceedings{15cddcdf0a664d4884c332e60337c6cf,
title = "Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion",
abstract = "Generation of object-boundary-matched accurate depth map is important to semi-automatic 2D to 3D conversion for high quality perception of converted 3D video. In this paper, we propose an accurate depth map generation scheme using an MRF-based initial contour tracking followed by a robust graph-cut-based contour refinement. To limit the propagation of potential tracking errors, key-frames as well as current and previous video frames are considered together in composing the graph cut seeds of the proposed contour refinement. In addition, the proposed scheme checks the ambiguity of seeds in foreground and background to eliminate the seeds, which may cause confusion in graph-cut segmentation, and thereby further improves the accuracy of contour tracking. Experiment results demonstrate the proposed scheme is especially useful when there is complex background or deformation of object, indicating the power of the proposed scheme in object-boundary-matched accurate depth map generation.",
author = "Junsoo Kim and Yoonsik Choe and Kim, {Yong Goo}",
year = "2011",
month = "10",
day = "24",
doi = "10.1109/PACRIM.2011.6032919",
language = "English",
isbn = "9781457702518",
series = "IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings",
pages = "358--363",
booktitle = "2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011",

}

Kim, J, Choe, Y & Kim, YG 2011, Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion. in 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011., 6032919, IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings, pp. 358-363, 13th IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011, Victoria, BC, Canada, 11/8/23. https://doi.org/10.1109/PACRIM.2011.6032919

Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion. / Kim, Junsoo; Choe, Yoonsik; Kim, Yong Goo.

2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011. 2011. p. 358-363 6032919 (IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings).

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

TY - GEN

T1 - Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion

AU - Kim, Junsoo

AU - Choe, Yoonsik

AU - Kim, Yong Goo

PY - 2011/10/24

Y1 - 2011/10/24

N2 - Generation of object-boundary-matched accurate depth map is important to semi-automatic 2D to 3D conversion for high quality perception of converted 3D video. In this paper, we propose an accurate depth map generation scheme using an MRF-based initial contour tracking followed by a robust graph-cut-based contour refinement. To limit the propagation of potential tracking errors, key-frames as well as current and previous video frames are considered together in composing the graph cut seeds of the proposed contour refinement. In addition, the proposed scheme checks the ambiguity of seeds in foreground and background to eliminate the seeds, which may cause confusion in graph-cut segmentation, and thereby further improves the accuracy of contour tracking. Experiment results demonstrate the proposed scheme is especially useful when there is complex background or deformation of object, indicating the power of the proposed scheme in object-boundary-matched accurate depth map generation.

AB - Generation of object-boundary-matched accurate depth map is important to semi-automatic 2D to 3D conversion for high quality perception of converted 3D video. In this paper, we propose an accurate depth map generation scheme using an MRF-based initial contour tracking followed by a robust graph-cut-based contour refinement. To limit the propagation of potential tracking errors, key-frames as well as current and previous video frames are considered together in composing the graph cut seeds of the proposed contour refinement. In addition, the proposed scheme checks the ambiguity of seeds in foreground and background to eliminate the seeds, which may cause confusion in graph-cut segmentation, and thereby further improves the accuracy of contour tracking. Experiment results demonstrate the proposed scheme is especially useful when there is complex background or deformation of object, indicating the power of the proposed scheme in object-boundary-matched accurate depth map generation.

UR - http://www.scopus.com/inward/record.url?scp=80054738657&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80054738657&partnerID=8YFLogxK

U2 - 10.1109/PACRIM.2011.6032919

DO - 10.1109/PACRIM.2011.6032919

M3 - Conference contribution

SN - 9781457702518

T3 - IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings

SP - 358

EP - 363

BT - 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011

ER -

Kim J, Choe Y, Kim YG. Robust MRF-based object tracking and graph-cut-based contour refinement for high quality 2D to 3D video conversion. In 2011 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, PACRIM 2011. 2011. p. 358-363. 6032919. (IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings). https://doi.org/10.1109/PACRIM.2011.6032919