Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy

M. Chen; X. Wei; Q. Yang; Qing Li; G. Wang; M.-H. Yang

doi:10.1109/TPAMI.2017.2717828

Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy

M. Chen, X. Wei, Q. Yang, Qing Li, G. Wang, M.-H. Yang

Research output: Journal article publication › Journal article › Academic research › peer-review

86 Citations (Scopus)

Abstract

© 1979-2012 IEEE. We propose a background subtraction algorithm using hierarchical superpixel segmentation, spanning trees and optical flow. First, we generate superpixel segmentation trees using a number of Gaussian Mixture Models (GMMs) by treating each GMM as one vertex to construct spanning trees. Next, we use the $M$-smoother to enhance the spatial consistency on the spanning trees and estimate optical flow to extend the $M$-smoother to the temporal domain. Experimental results on synthetic and real-world benchmark datasets show that the proposed algorithm performs favorably for background subtraction in videos against the state-of-the-art methods in spite of frequent and sudden changes of pixel values.

Original language	English
Pages (from-to)	1518-1525
Number of pages	8
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	40
Issue number	6
DOIs	https://doi.org/10.1109/TPAMI.2017.2717828
Publication status	Published - 1 Jun 2018
Externally published	Yes

Keywords

Background modeling
minimum spanning tree
optical flow
superpixel hierarchy
tracking

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Computational Theory and Mathematics
Artificial Intelligence
Applied Mathematics

More information

10.1109/TPAMI.2017.2717828

Cite this

@article{4cbfb24b4734499a9fb95565c688f608,

title = "Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy",

abstract = "{\textcopyright} 1979-2012 IEEE. We propose a background subtraction algorithm using hierarchical superpixel segmentation, spanning trees and optical flow. First, we generate superpixel segmentation trees using a number of Gaussian Mixture Models (GMMs) by treating each GMM as one vertex to construct spanning trees. Next, we use the $M$-smoother to enhance the spatial consistency on the spanning trees and estimate optical flow to extend the $M$-smoother to the temporal domain. Experimental results on synthetic and real-world benchmark datasets show that the proposed algorithm performs favorably for background subtraction in videos against the state-of-the-art methods in spite of frequent and sudden changes of pixel values.",

keywords = "Background modeling, minimum spanning tree, optical flow, superpixel hierarchy, tracking",

author = "M. Chen and X. Wei and Q. Yang and Qing Li and G. Wang and M.-H. Yang",

year = "2018",

month = jun,

day = "1",

doi = "10.1109/TPAMI.2017.2717828",

language = "English",

volume = "40",

pages = "1518--1525",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "6",

}

TY - JOUR

T1 - Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy

AU - Chen, M.

AU - Wei, X.

AU - Yang, Q.

AU - Li, Qing

AU - Wang, G.

AU - Yang, M.-H.

PY - 2018/6/1

Y1 - 2018/6/1

N2 - © 1979-2012 IEEE. We propose a background subtraction algorithm using hierarchical superpixel segmentation, spanning trees and optical flow. First, we generate superpixel segmentation trees using a number of Gaussian Mixture Models (GMMs) by treating each GMM as one vertex to construct spanning trees. Next, we use the $M$-smoother to enhance the spatial consistency on the spanning trees and estimate optical flow to extend the $M$-smoother to the temporal domain. Experimental results on synthetic and real-world benchmark datasets show that the proposed algorithm performs favorably for background subtraction in videos against the state-of-the-art methods in spite of frequent and sudden changes of pixel values.

AB - © 1979-2012 IEEE. We propose a background subtraction algorithm using hierarchical superpixel segmentation, spanning trees and optical flow. First, we generate superpixel segmentation trees using a number of Gaussian Mixture Models (GMMs) by treating each GMM as one vertex to construct spanning trees. Next, we use the $M$-smoother to enhance the spatial consistency on the spanning trees and estimate optical flow to extend the $M$-smoother to the temporal domain. Experimental results on synthetic and real-world benchmark datasets show that the proposed algorithm performs favorably for background subtraction in videos against the state-of-the-art methods in spite of frequent and sudden changes of pixel values.

KW - Background modeling

KW - minimum spanning tree

KW - optical flow

KW - superpixel hierarchy

KW - tracking

UR - http://www.scopus.com/inward/record.url?scp=85021792303&partnerID=8YFLogxK

U2 - 10.1109/TPAMI.2017.2717828

DO - 10.1109/TPAMI.2017.2717828

M3 - Journal article

SN - 0162-8828

VL - 40

SP - 1518

EP - 1525

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 6

ER -

Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy

Abstract

Keywords

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this