Drift robust non-rigid optical flow enhancement for long sequences

Wenbin Li, Darren Cosker, Matthew Brown

Research output: Contribution to journalArticle

5 Citations (Scopus)
85 Downloads (Pure)

Abstract

It is hard to densely track a nonrigid object in long term, which is a fundamental research issue in the computer vision community. This task often relies on estimating pairwise correspondences between images over time where the error is accumulated and leads to a drift. In this paper, we introduce a novel optimisation framework with an Anchor Patch constraint. It is supposed to significantly reduce overall errors given long sequences containing nonrigidly deformable objects. Our framework can be applied to any dense tracking algorithm, e.g. optical flow. We demonstrate the success of our approach by showing significant error reduction on 6 popular optical flow algorithms applied to a range of realworld nonrigid benchmarks. We also provide quantitative analysis of our approach given synthetic occlusions and image noise.
Original languageEnglish
Pages (from-to)2583-2595
Number of pages13
JournalJournal of Intelligent and Fuzzy Systems
Volume31
Issue number5
DOIs
Publication statusPublished - 13 Oct 2016

Fingerprint

Optical flows
Optical Flow
Enhancement
Deformable Objects
Error Reduction
Quantitative Analysis
Occlusion
Computer Vision
Patch
Pairwise
Correspondence
Benchmark
Anchors
Computer vision
Optimization
Term
Range of data
Demonstrate
Chemical analysis
Framework

Cite this

Drift robust non-rigid optical flow enhancement for long sequences. / Li, Wenbin; Cosker, Darren; Brown, Matthew.

In: Journal of Intelligent and Fuzzy Systems, Vol. 31, No. 5, 13.10.2016, p. 2583-2595.

Research output: Contribution to journalArticle

@article{08ef047e35a8494fb4fc9397cd2e49e2,
title = "Drift robust non-rigid optical flow enhancement for long sequences",
abstract = "It is hard to densely track a nonrigid object in long term, which is a fundamental research issue in the computer vision community. This task often relies on estimating pairwise correspondences between images over time where the error is accumulated and leads to a drift. In this paper, we introduce a novel optimisation framework with an Anchor Patch constraint. It is supposed to significantly reduce overall errors given long sequences containing nonrigidly deformable objects. Our framework can be applied to any dense tracking algorithm, e.g. optical flow. We demonstrate the success of our approach by showing significant error reduction on 6 popular optical flow algorithms applied to a range of realworld nonrigid benchmarks. We also provide quantitative analysis of our approach given synthetic occlusions and image noise.",
author = "Wenbin Li and Darren Cosker and Matthew Brown",
year = "2016",
month = "10",
day = "13",
doi = "10.3233/JIFS-169099",
language = "English",
volume = "31",
pages = "2583--2595",
journal = "Journal of Intelligent and Fuzzy Systems",
issn = "1064-1246",
publisher = "IOS Press",
number = "5",

}

TY - JOUR

T1 - Drift robust non-rigid optical flow enhancement for long sequences

AU - Li, Wenbin

AU - Cosker, Darren

AU - Brown, Matthew

PY - 2016/10/13

Y1 - 2016/10/13

N2 - It is hard to densely track a nonrigid object in long term, which is a fundamental research issue in the computer vision community. This task often relies on estimating pairwise correspondences between images over time where the error is accumulated and leads to a drift. In this paper, we introduce a novel optimisation framework with an Anchor Patch constraint. It is supposed to significantly reduce overall errors given long sequences containing nonrigidly deformable objects. Our framework can be applied to any dense tracking algorithm, e.g. optical flow. We demonstrate the success of our approach by showing significant error reduction on 6 popular optical flow algorithms applied to a range of realworld nonrigid benchmarks. We also provide quantitative analysis of our approach given synthetic occlusions and image noise.

AB - It is hard to densely track a nonrigid object in long term, which is a fundamental research issue in the computer vision community. This task often relies on estimating pairwise correspondences between images over time where the error is accumulated and leads to a drift. In this paper, we introduce a novel optimisation framework with an Anchor Patch constraint. It is supposed to significantly reduce overall errors given long sequences containing nonrigidly deformable objects. Our framework can be applied to any dense tracking algorithm, e.g. optical flow. We demonstrate the success of our approach by showing significant error reduction on 6 popular optical flow algorithms applied to a range of realworld nonrigid benchmarks. We also provide quantitative analysis of our approach given synthetic occlusions and image noise.

UR - http://dx.doi.org/10.3233/JIFS-169099

U2 - 10.3233/JIFS-169099

DO - 10.3233/JIFS-169099

M3 - Article

VL - 31

SP - 2583

EP - 2595

JO - Journal of Intelligent and Fuzzy Systems

JF - Journal of Intelligent and Fuzzy Systems

SN - 1064-1246

IS - 5

ER -