Abstract

Dense motion estimations obtained from optical flow techniques play a significant role in many image processing and computer vision tasks. Remarkable progress has been made in both theory and its application in practice. In this paper, we provide a systematic review of recent optical flow techniques with a focus on the variational method and approaches based on Convolutional Neural Networks (CNNs). These two categories have led to state-of-the-art performance. We discuss recent modifications and extensions of the original model, and highlight remaining challenges. For the first time, we provide an overview of recent CNN-based optical flow methods and discuss their potential and current limitations.

Original languageEnglish (US)
Pages (from-to)9-24
Number of pages16
JournalSignal Processing: Image Communication
Volume72
DOIs
StatePublished - Mar 1 2019

Fingerprint

Optical flows
Neural networks
Motion estimation
Computer vision
Image processing

Keywords

  • Challenges
  • CNN-based method
  • Evaluation measures
  • Optical flow
  • Variational method

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering

Cite this

A survey of variational and CNN-based optical flow techniques. / Tu, Zhigang; Xie, Wei; Zhang, Dejun; Poppe, Ronald; Veltkamp, Remco C.; Li, Baoxin; Yuan, Junsong.

In: Signal Processing: Image Communication, Vol. 72, 01.03.2019, p. 9-24.

Research output: Contribution to journalArticle

Tu, Zhigang ; Xie, Wei ; Zhang, Dejun ; Poppe, Ronald ; Veltkamp, Remco C. ; Li, Baoxin ; Yuan, Junsong. / A survey of variational and CNN-based optical flow techniques. In: Signal Processing: Image Communication. 2019 ; Vol. 72. pp. 9-24.
@article{284d151d5de04959a24b143558753303,
title = "A survey of variational and CNN-based optical flow techniques",
abstract = "Dense motion estimations obtained from optical flow techniques play a significant role in many image processing and computer vision tasks. Remarkable progress has been made in both theory and its application in practice. In this paper, we provide a systematic review of recent optical flow techniques with a focus on the variational method and approaches based on Convolutional Neural Networks (CNNs). These two categories have led to state-of-the-art performance. We discuss recent modifications and extensions of the original model, and highlight remaining challenges. For the first time, we provide an overview of recent CNN-based optical flow methods and discuss their potential and current limitations.",
keywords = "Challenges, CNN-based method, Evaluation measures, Optical flow, Variational method",
author = "Zhigang Tu and Wei Xie and Dejun Zhang and Ronald Poppe and Veltkamp, {Remco C.} and Baoxin Li and Junsong Yuan",
year = "2019",
month = "3",
day = "1",
doi = "10.1016/j.image.2018.12.002",
language = "English (US)",
volume = "72",
pages = "9--24",
journal = "Signal Processing: Image Communication",
issn = "0923-5965",
publisher = "Elsevier",

}

TY - JOUR

T1 - A survey of variational and CNN-based optical flow techniques

AU - Tu, Zhigang

AU - Xie, Wei

AU - Zhang, Dejun

AU - Poppe, Ronald

AU - Veltkamp, Remco C.

AU - Li, Baoxin

AU - Yuan, Junsong

PY - 2019/3/1

Y1 - 2019/3/1

N2 - Dense motion estimations obtained from optical flow techniques play a significant role in many image processing and computer vision tasks. Remarkable progress has been made in both theory and its application in practice. In this paper, we provide a systematic review of recent optical flow techniques with a focus on the variational method and approaches based on Convolutional Neural Networks (CNNs). These two categories have led to state-of-the-art performance. We discuss recent modifications and extensions of the original model, and highlight remaining challenges. For the first time, we provide an overview of recent CNN-based optical flow methods and discuss their potential and current limitations.

AB - Dense motion estimations obtained from optical flow techniques play a significant role in many image processing and computer vision tasks. Remarkable progress has been made in both theory and its application in practice. In this paper, we provide a systematic review of recent optical flow techniques with a focus on the variational method and approaches based on Convolutional Neural Networks (CNNs). These two categories have led to state-of-the-art performance. We discuss recent modifications and extensions of the original model, and highlight remaining challenges. For the first time, we provide an overview of recent CNN-based optical flow methods and discuss their potential and current limitations.

KW - Challenges

KW - CNN-based method

KW - Evaluation measures

KW - Optical flow

KW - Variational method

UR - http://www.scopus.com/inward/record.url?scp=85058447980&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058447980&partnerID=8YFLogxK

U2 - 10.1016/j.image.2018.12.002

DO - 10.1016/j.image.2018.12.002

M3 - Article

VL - 72

SP - 9

EP - 24

JO - Signal Processing: Image Communication

JF - Signal Processing: Image Communication

SN - 0923-5965

ER -