Optical Flow Estimation using Fourier-Mellin Transform

Huy Tho Ho, Roland Goecke

Research output: A Conference proceeding or a Chapter in BookConference contribution

20 Citations (Scopus)

Abstract

In this paper, we propose a novel method of computing the optical flow using the Fourier Mellin Transform (FMT). Each image in a sequence is divided into a regular grid of patches and the optical flow is estimated by calculating the phase correlation of each pair of co-sited patches using the FMT. By applying the FMT in calculating the phase correlation, we are able to estimate not only the pure translation, as limited in the case of the basic phase correlation techniques, but also the scale and rotation motion of image patches, i.e. full similarity transforms. Moreover, the motion parameters of each patch can be estimated to sub-pixel accuracy based on a recently proposed algorithm that uses a 2D esinc function in fitting the data from the phase correlation output. We also improve the estimation of the optical flow by presenting a method of smoothing the field by using a vector weighted average filter. Finally, experimental results, using publicly available data sets are presented, demonstrating the accuracy and improvements of our method over previous optical flow methods.
Original languageEnglish
Title of host publicationIEEE CVPR 2008
EditorsOmar Javed
Place of PublicationAustralia
PublisherIEEE
Pages1-23
Number of pages23
DOIs
Publication statusPublished - 2008
Externally publishedYes
EventIEEE CVPR2008 - Anchorage, United States
Duration: 24 Jun 200826 Jun 2008

Conference

ConferenceIEEE CVPR2008
CountryUnited States
CityAnchorage
Period24/06/0826/06/08

Fingerprint

Optical flows
Fourier transforms
Pixels

Cite this

Ho, H. T., & Goecke, R. (2008). Optical Flow Estimation using Fourier-Mellin Transform. In O. Javed (Ed.), IEEE CVPR 2008 (pp. 1-23). Australia: IEEE. https://doi.org/10.1109/CVPR.2008.4587553
Ho, Huy Tho ; Goecke, Roland. / Optical Flow Estimation using Fourier-Mellin Transform. IEEE CVPR 2008. editor / Omar Javed. Australia : IEEE, 2008. pp. 1-23
@inproceedings{6c60af0d9fe345e6829b8a7d72d8656b,
title = "Optical Flow Estimation using Fourier-Mellin Transform",
abstract = "In this paper, we propose a novel method of computing the optical flow using the Fourier Mellin Transform (FMT). Each image in a sequence is divided into a regular grid of patches and the optical flow is estimated by calculating the phase correlation of each pair of co-sited patches using the FMT. By applying the FMT in calculating the phase correlation, we are able to estimate not only the pure translation, as limited in the case of the basic phase correlation techniques, but also the scale and rotation motion of image patches, i.e. full similarity transforms. Moreover, the motion parameters of each patch can be estimated to sub-pixel accuracy based on a recently proposed algorithm that uses a 2D esinc function in fitting the data from the phase correlation output. We also improve the estimation of the optical flow by presenting a method of smoothing the field by using a vector weighted average filter. Finally, experimental results, using publicly available data sets are presented, demonstrating the accuracy and improvements of our method over previous optical flow methods.",
author = "Ho, {Huy Tho} and Roland Goecke",
year = "2008",
doi = "10.1109/CVPR.2008.4587553",
language = "English",
pages = "1--23",
editor = "Omar Javed",
booktitle = "IEEE CVPR 2008",
publisher = "IEEE",

}

Ho, HT & Goecke, R 2008, Optical Flow Estimation using Fourier-Mellin Transform. in O Javed (ed.), IEEE CVPR 2008. IEEE, Australia, pp. 1-23, IEEE CVPR2008, Anchorage, United States, 24/06/08. https://doi.org/10.1109/CVPR.2008.4587553

Optical Flow Estimation using Fourier-Mellin Transform. / Ho, Huy Tho; Goecke, Roland.

IEEE CVPR 2008. ed. / Omar Javed. Australia : IEEE, 2008. p. 1-23.

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Optical Flow Estimation using Fourier-Mellin Transform

AU - Ho, Huy Tho

AU - Goecke, Roland

PY - 2008

Y1 - 2008

N2 - In this paper, we propose a novel method of computing the optical flow using the Fourier Mellin Transform (FMT). Each image in a sequence is divided into a regular grid of patches and the optical flow is estimated by calculating the phase correlation of each pair of co-sited patches using the FMT. By applying the FMT in calculating the phase correlation, we are able to estimate not only the pure translation, as limited in the case of the basic phase correlation techniques, but also the scale and rotation motion of image patches, i.e. full similarity transforms. Moreover, the motion parameters of each patch can be estimated to sub-pixel accuracy based on a recently proposed algorithm that uses a 2D esinc function in fitting the data from the phase correlation output. We also improve the estimation of the optical flow by presenting a method of smoothing the field by using a vector weighted average filter. Finally, experimental results, using publicly available data sets are presented, demonstrating the accuracy and improvements of our method over previous optical flow methods.

AB - In this paper, we propose a novel method of computing the optical flow using the Fourier Mellin Transform (FMT). Each image in a sequence is divided into a regular grid of patches and the optical flow is estimated by calculating the phase correlation of each pair of co-sited patches using the FMT. By applying the FMT in calculating the phase correlation, we are able to estimate not only the pure translation, as limited in the case of the basic phase correlation techniques, but also the scale and rotation motion of image patches, i.e. full similarity transforms. Moreover, the motion parameters of each patch can be estimated to sub-pixel accuracy based on a recently proposed algorithm that uses a 2D esinc function in fitting the data from the phase correlation output. We also improve the estimation of the optical flow by presenting a method of smoothing the field by using a vector weighted average filter. Finally, experimental results, using publicly available data sets are presented, demonstrating the accuracy and improvements of our method over previous optical flow methods.

U2 - 10.1109/CVPR.2008.4587553

DO - 10.1109/CVPR.2008.4587553

M3 - Conference contribution

SP - 1

EP - 23

BT - IEEE CVPR 2008

A2 - Javed, Omar

PB - IEEE

CY - Australia

ER -

Ho HT, Goecke R. Optical Flow Estimation using Fourier-Mellin Transform. In Javed O, editor, IEEE CVPR 2008. Australia: IEEE. 2008. p. 1-23 https://doi.org/10.1109/CVPR.2008.4587553