Subspace extension to phase correlation approach for fast image registration

Jinchang Ren, T. Vlachos, J. Jiang

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

Abstract

A novel extension of phase correlation to subspace correlation is proposed, in which 2-D translation is decomposed into two I-D motions thus only 1-D Fourier transform is used to estimate the corresponding motion. In each subspace, the first two highest peaks from I-D correlation are linearly interpolated for subpixel accuracy. Experimental results have shown both the robustness and accuracy of our method.
Original languageEnglish
Title of host publication2007 IEEE International conference on image processing, Volumes 1-7
Place of PublicationNew York
PublisherIEEE
Pages481-484
Number of pages4
ISBN (Print)9781424414369
Publication statusPublished - 2007
Event2007 14th IEEE International Conference on Image Processing - Hyatt Regency San Antonio, San Antonio, United States
Duration: 16 Sep 200719 Sep 2007

Publication series

NameIEEE International Conference on Image Processing (ICIP)
PublisherIEEE
ISSN (Print)1522-4880

Conference

Conference2007 14th IEEE International Conference on Image Processing
Abbreviated titleICIP 2007
CountryUnited States
CitySan Antonio
Period16/09/0719/09/07

    Fingerprint

Keywords

  • fourier transform
  • subspace projection
  • subpixel alignment
  • image registration
  • phase correlation

Cite this

Ren, J., Vlachos, T., & Jiang , J. (2007). Subspace extension to phase correlation approach for fast image registration. In 2007 IEEE International conference on image processing, Volumes 1-7 (pp. 481-484). (IEEE International Conference on Image Processing (ICIP) ). New York: IEEE.