Total variation and Rank-1 constraint RPCA for background subtraction

Jize Xue, Yongqiang Zhao, Wenzhi Liao, Jonathan Cheung-Wai Chan

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Background subtraction (BS) in video sequences is a main research field, and the aim is to separate moving objects in the foreground from stationary background. Using the framework of schemes-based robust principal component analysis (RPCA), we propose a novel BS method employing the more refined prior representations for the static and dynamic components of the video sequences. Specifically, the rank-1 constraint is exploited to describe the strong low-rank property of background layer (temporal correlation of static component), and 3-D total variation measure and L 1 norm are used to model the spatial-temporal smoothness of foreground layer and sparseness of noise (dynamic component). This method introduces rank-1, smooth, and sparse properties into the RPCA framework for BS task, and it is dubbed TR1-RPCA. In addition, an efficient algorithm based on the alternating direction method of multipliers is designed to solve the proposed BS model. Extensive experiments on simulated and real videos demonstrate the superiority of the proposed method.
LanguageEnglish
Pages49955-49966
Number of pages12
JournalIEEE Access
Volume6
DOIs
Publication statusAccepted/In press - 25 Aug 2018

Fingerprint

principal components analysis
subtraction
Principal component analysis
principal component analysis
multipliers
norms
method
Experiments
video
experiment

Keywords

  • background subtraction
  • total variation
  • rank-1 property
  • robust principal component analysis
  • spatial-temporal correlations
  • task analysis
  • video sequences
  • heuristic algorithms

Cite this

Xue, Jize ; Zhao, Yongqiang ; Liao, Wenzhi ; Chan, Jonathan Cheung-Wai. / Total variation and Rank-1 constraint RPCA for background subtraction. In: IEEE Access. 2018 ; Vol. 6. pp. 49955-49966.
@article{5fc54c3c2e434694b160c4527f08bc44,
title = "Total variation and Rank-1 constraint RPCA for background subtraction",
abstract = "Background subtraction (BS) in video sequences is a main research field, and the aim is to separate moving objects in the foreground from stationary background. Using the framework of schemes-based robust principal component analysis (RPCA), we propose a novel BS method employing the more refined prior representations for the static and dynamic components of the video sequences. Specifically, the rank-1 constraint is exploited to describe the strong low-rank property of background layer (temporal correlation of static component), and 3-D total variation measure and L 1 norm are used to model the spatial-temporal smoothness of foreground layer and sparseness of noise (dynamic component). This method introduces rank-1, smooth, and sparse properties into the RPCA framework for BS task, and it is dubbed TR1-RPCA. In addition, an efficient algorithm based on the alternating direction method of multipliers is designed to solve the proposed BS model. Extensive experiments on simulated and real videos demonstrate the superiority of the proposed method.",
keywords = "background subtraction, total variation, rank-1 property, robust principal component analysis, spatial-temporal correlations, task analysis, video sequences, heuristic algorithms",
author = "Jize Xue and Yongqiang Zhao and Wenzhi Liao and Chan, {Jonathan Cheung-Wai}",
note = "(c) 2018 IEEE.",
year = "2018",
month = "8",
day = "25",
doi = "10.1109/ACCESS.2018.2868731",
language = "English",
volume = "6",
pages = "49955--49966",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "IEEE",

}

Total variation and Rank-1 constraint RPCA for background subtraction. / Xue, Jize; Zhao, Yongqiang; Liao, Wenzhi; Chan, Jonathan Cheung-Wai.

In: IEEE Access, Vol. 6, 25.08.2018, p. 49955-49966.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Total variation and Rank-1 constraint RPCA for background subtraction

AU - Xue, Jize

AU - Zhao, Yongqiang

AU - Liao, Wenzhi

AU - Chan, Jonathan Cheung-Wai

N1 - (c) 2018 IEEE.

PY - 2018/8/25

Y1 - 2018/8/25

N2 - Background subtraction (BS) in video sequences is a main research field, and the aim is to separate moving objects in the foreground from stationary background. Using the framework of schemes-based robust principal component analysis (RPCA), we propose a novel BS method employing the more refined prior representations for the static and dynamic components of the video sequences. Specifically, the rank-1 constraint is exploited to describe the strong low-rank property of background layer (temporal correlation of static component), and 3-D total variation measure and L 1 norm are used to model the spatial-temporal smoothness of foreground layer and sparseness of noise (dynamic component). This method introduces rank-1, smooth, and sparse properties into the RPCA framework for BS task, and it is dubbed TR1-RPCA. In addition, an efficient algorithm based on the alternating direction method of multipliers is designed to solve the proposed BS model. Extensive experiments on simulated and real videos demonstrate the superiority of the proposed method.

AB - Background subtraction (BS) in video sequences is a main research field, and the aim is to separate moving objects in the foreground from stationary background. Using the framework of schemes-based robust principal component analysis (RPCA), we propose a novel BS method employing the more refined prior representations for the static and dynamic components of the video sequences. Specifically, the rank-1 constraint is exploited to describe the strong low-rank property of background layer (temporal correlation of static component), and 3-D total variation measure and L 1 norm are used to model the spatial-temporal smoothness of foreground layer and sparseness of noise (dynamic component). This method introduces rank-1, smooth, and sparse properties into the RPCA framework for BS task, and it is dubbed TR1-RPCA. In addition, an efficient algorithm based on the alternating direction method of multipliers is designed to solve the proposed BS model. Extensive experiments on simulated and real videos demonstrate the superiority of the proposed method.

KW - background subtraction

KW - total variation

KW - rank-1 property

KW - robust principal component analysis

KW - spatial-temporal correlations

KW - task analysis

KW - video sequences

KW - heuristic algorithms

U2 - 10.1109/ACCESS.2018.2868731

DO - 10.1109/ACCESS.2018.2868731

M3 - Article

VL - 6

SP - 49955

EP - 49966

JO - IEEE Access

T2 - IEEE Access

JF - IEEE Access

SN - 2169-3536

ER -