A Frequency Domain Performance Analysis of Horn and Schunck's Optical Flow Algorithm for Deformable Motion

Thomas S. Denney, Jerry Ladd Prince

Research output: Contribution to journalArticle

Abstract

A frequency domain performance analysis of Horn and Schanck's optical flow (HSOF) algorithm for estimation of deformable motion is presented. Noise sources in the algorithm are modeled using the discrete Fourier transform of the brightness pattern. This noise model along with the estimation error covariance function derived in previous work is used to derive an expression for the expected performance of the optical flow estimate that is valid for an arbitrary discrete brightness pattern. Simulation results are presented that demonstrate the validity of our methods and show that HSOF is more accurate that the optical flow estimate of Anandan for certain low-frequency patterns.

Original languageEnglish (US)
Pages (from-to)1324-1327
Number of pages4
JournalIEEE Transactions on Image Processing
Volume4
Issue number9
DOIs
StatePublished - 1995

Fingerprint

Domain Analysis
Optical flows
Optical Flow
Performance Analysis
Frequency Domain
Motion
Brightness
Luminance
Covariance Function
Discrete Fourier transform
Error function
Estimation Error
Discrete Fourier transforms
Estimate
Error analysis
Low Frequency
Valid
Arbitrary
Demonstrate
Simulation

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software
  • Computational Theory and Mathematics
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Theoretical Computer Science

Cite this

A Frequency Domain Performance Analysis of Horn and Schunck's Optical Flow Algorithm for Deformable Motion. / Denney, Thomas S.; Prince, Jerry Ladd.

In: IEEE Transactions on Image Processing, Vol. 4, No. 9, 1995, p. 1324-1327.

Research output: Contribution to journalArticle

@article{d19ff95598f74556821c3c24a845bec4,
title = "A Frequency Domain Performance Analysis of Horn and Schunck's Optical Flow Algorithm for Deformable Motion",
abstract = "A frequency domain performance analysis of Horn and Schanck's optical flow (HSOF) algorithm for estimation of deformable motion is presented. Noise sources in the algorithm are modeled using the discrete Fourier transform of the brightness pattern. This noise model along with the estimation error covariance function derived in previous work is used to derive an expression for the expected performance of the optical flow estimate that is valid for an arbitrary discrete brightness pattern. Simulation results are presented that demonstrate the validity of our methods and show that HSOF is more accurate that the optical flow estimate of Anandan for certain low-frequency patterns.",
author = "Denney, {Thomas S.} and Prince, {Jerry Ladd}",
year = "1995",
doi = "10.1109/83.413178",
language = "English (US)",
volume = "4",
pages = "1324--1327",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "9",

}

TY - JOUR

T1 - A Frequency Domain Performance Analysis of Horn and Schunck's Optical Flow Algorithm for Deformable Motion

AU - Denney, Thomas S.

AU - Prince, Jerry Ladd

PY - 1995

Y1 - 1995

N2 - A frequency domain performance analysis of Horn and Schanck's optical flow (HSOF) algorithm for estimation of deformable motion is presented. Noise sources in the algorithm are modeled using the discrete Fourier transform of the brightness pattern. This noise model along with the estimation error covariance function derived in previous work is used to derive an expression for the expected performance of the optical flow estimate that is valid for an arbitrary discrete brightness pattern. Simulation results are presented that demonstrate the validity of our methods and show that HSOF is more accurate that the optical flow estimate of Anandan for certain low-frequency patterns.

AB - A frequency domain performance analysis of Horn and Schanck's optical flow (HSOF) algorithm for estimation of deformable motion is presented. Noise sources in the algorithm are modeled using the discrete Fourier transform of the brightness pattern. This noise model along with the estimation error covariance function derived in previous work is used to derive an expression for the expected performance of the optical flow estimate that is valid for an arbitrary discrete brightness pattern. Simulation results are presented that demonstrate the validity of our methods and show that HSOF is more accurate that the optical flow estimate of Anandan for certain low-frequency patterns.

UR - http://www.scopus.com/inward/record.url?scp=0029377657&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029377657&partnerID=8YFLogxK

U2 - 10.1109/83.413178

DO - 10.1109/83.413178

M3 - Article

C2 - 18292029

AN - SCOPUS:0029377657

VL - 4

SP - 1324

EP - 1327

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 9

ER -