An EM based approach for motion segmentation of video sequence

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

Abstract

Motions are important features for robot vision as we live in a dynamic world. Detecting moving objects is crucial for mobile robots and computer vision systems. This paper investigates an architecture for the segmentation of moving objects from image sequences. Objects are represented as groups of SIFT feature points. Instead of tracking the feature points over a sequence of frames, the movements of feature points between two successive frames are used. The segmentation of motions of each pair of frames is based on the expectation-maximization algorithm. The segmentation algorithm is iteratively applied over all frames of the sequence and the results are combined using Bayesian update.
Original languageEnglish
Title of host publication International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG)
EditorsZhigeng Pan, Vaclav Skala
PublisherVaclav Skala Union Agency
Pages61-69
Number of pages9
ISBN (Electronic)978-80-86943-57-2
Publication statusPublished - 2016

Publication series

SeriesComputer Science Research Notes
Volume2601
ISSN2464-4617

Cite this

Zhao, W., & Roos, N. (2016). An EM based approach for motion segmentation of video sequence. In Z. Pan, & V. Skala (Eds.), International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG) (pp. 61-69). Vaclav Skala Union Agency. Computer Science Research Notes, Vol.. 2601
Zhao, Wei ; Roos, Nico. / An EM based approach for motion segmentation of video sequence. International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG). editor / Zhigeng Pan ; Vaclav Skala. Vaclav Skala Union Agency, 2016. pp. 61-69 (Computer Science Research Notes, Vol. 2601).
@inproceedings{bffb46a7be9040299be5b009fc9cef0b,
title = "An EM based approach for motion segmentation of video sequence",
abstract = "Motions are important features for robot vision as we live in a dynamic world. Detecting moving objects is crucial for mobile robots and computer vision systems. This paper investigates an architecture for the segmentation of moving objects from image sequences. Objects are represented as groups of SIFT feature points. Instead of tracking the feature points over a sequence of frames, the movements of feature points between two successive frames are used. The segmentation of motions of each pair of frames is based on the expectation-maximization algorithm. The segmentation algorithm is iteratively applied over all frames of the sequence and the results are combined using Bayesian update.",
author = "Wei Zhao and Nico Roos",
year = "2016",
language = "English",
series = "Computer Science Research Notes",
pages = "61--69",
editor = "Zhigeng Pan and Vaclav Skala",
booktitle = "International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG)",
publisher = "Vaclav Skala Union Agency",
address = "Czech Republic",

}

Zhao, W & Roos, N 2016, An EM based approach for motion segmentation of video sequence. in Z Pan & V Skala (eds), International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG). Vaclav Skala Union Agency, Computer Science Research Notes, vol. 2601, pp. 61-69.

An EM based approach for motion segmentation of video sequence. / Zhao, Wei; Roos, Nico.

International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG). ed. / Zhigeng Pan; Vaclav Skala. Vaclav Skala Union Agency, 2016. p. 61-69 (Computer Science Research Notes, Vol. 2601).

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

TY - GEN

T1 - An EM based approach for motion segmentation of video sequence

AU - Zhao, Wei

AU - Roos, Nico

PY - 2016

Y1 - 2016

N2 - Motions are important features for robot vision as we live in a dynamic world. Detecting moving objects is crucial for mobile robots and computer vision systems. This paper investigates an architecture for the segmentation of moving objects from image sequences. Objects are represented as groups of SIFT feature points. Instead of tracking the feature points over a sequence of frames, the movements of feature points between two successive frames are used. The segmentation of motions of each pair of frames is based on the expectation-maximization algorithm. The segmentation algorithm is iteratively applied over all frames of the sequence and the results are combined using Bayesian update.

AB - Motions are important features for robot vision as we live in a dynamic world. Detecting moving objects is crucial for mobile robots and computer vision systems. This paper investigates an architecture for the segmentation of moving objects from image sequences. Objects are represented as groups of SIFT feature points. Instead of tracking the feature points over a sequence of frames, the movements of feature points between two successive frames are used. The segmentation of motions of each pair of frames is based on the expectation-maximization algorithm. The segmentation algorithm is iteratively applied over all frames of the sequence and the results are combined using Bayesian update.

M3 - Conference article in proceeding

T3 - Computer Science Research Notes

SP - 61

EP - 69

BT - International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG)

A2 - Pan, Zhigeng

A2 - Skala, Vaclav

PB - Vaclav Skala Union Agency

ER -

Zhao W, Roos N. An EM based approach for motion segmentation of video sequence. In Pan Z, Skala V, editors, International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG). Vaclav Skala Union Agency. 2016. p. 61-69. (Computer Science Research Notes, Vol. 2601).