Sub-pixel SAR image registration through parabolic interpolation of the 2D cross-correlation

Luca Pallotta, Gaetano Giunta, Carmine Clemente

Research output: Contribution to journalArticle

1 Downloads (Pure)

Abstract

In this paper the problem of Synthetic Aperture Radar (SAR) images coregistration is considered. In particular, a novel algorithm aimed at achieving a fine sub-pixel coregistration accuracy is developed. The procedure is based on the parabolic interpolation of the 2D cross-correlation computed between the two SAR images to be aligned. More precisely, from the Two- Dimensional (2D) cross-correlation, a neighborhood of its peak value is extracted and the interpolation of both the 2D paraboloid and the two alternative 1D parabolas are computed to provide the finer misregistration estimation with a sub-pixel accuracy. The main advantage of the proposed framework is that the overall computational burden is only due to the 2D cross-correlation estimation since the parabolic interpolation is calculated with a closed-form expression. The results obtained on real recorded Unmanned Aerial Vehicle (UAV) SAR data highlight the effectiveness of the proposed approach as well as its capabilities to provide some benefits with respect to other available strategy.
Original languageEnglish
Number of pages14
JournalIEEE Transactions on Geoscience and Remote Sensing
Publication statusAccepted/In press - 18 Dec 2019

Fingerprint

Image registration
Synthetic aperture radar
interpolation
Interpolation
pixel
synthetic aperture radar
Pixels
Unmanned aerial vehicles (UAV)
image registration

Keywords

  • synthetic aperture radar
  • SAR coregistration
  • 2D cross-correlation
  • parabolic interpolation
  • sub-pixel coregistration
  • UAV SAR

Cite this

@article{00a80deacaee42a0a1eb7e592a40d590,
title = "Sub-pixel SAR image registration through parabolic interpolation of the 2D cross-correlation",
abstract = "In this paper the problem of Synthetic Aperture Radar (SAR) images coregistration is considered. In particular, a novel algorithm aimed at achieving a fine sub-pixel coregistration accuracy is developed. The procedure is based on the parabolic interpolation of the 2D cross-correlation computed between the two SAR images to be aligned. More precisely, from the Two- Dimensional (2D) cross-correlation, a neighborhood of its peak value is extracted and the interpolation of both the 2D paraboloid and the two alternative 1D parabolas are computed to provide the finer misregistration estimation with a sub-pixel accuracy. The main advantage of the proposed framework is that the overall computational burden is only due to the 2D cross-correlation estimation since the parabolic interpolation is calculated with a closed-form expression. The results obtained on real recorded Unmanned Aerial Vehicle (UAV) SAR data highlight the effectiveness of the proposed approach as well as its capabilities to provide some benefits with respect to other available strategy.",
keywords = "synthetic aperture radar, SAR coregistration, 2D cross-correlation, parabolic interpolation, sub-pixel coregistration, UAV SAR",
author = "Luca Pallotta and Gaetano Giunta and Carmine Clemente",
note = "{\circledC} 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.",
year = "2019",
month = "12",
day = "18",
language = "English",
journal = "IEEE Transactions on Geoscience and Remote Sensing",
issn = "0196-2892",
publisher = "IEEE",

}

TY - JOUR

T1 - Sub-pixel SAR image registration through parabolic interpolation of the 2D cross-correlation

AU - Pallotta, Luca

AU - Giunta, Gaetano

AU - Clemente, Carmine

N1 - © 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2019/12/18

Y1 - 2019/12/18

N2 - In this paper the problem of Synthetic Aperture Radar (SAR) images coregistration is considered. In particular, a novel algorithm aimed at achieving a fine sub-pixel coregistration accuracy is developed. The procedure is based on the parabolic interpolation of the 2D cross-correlation computed between the two SAR images to be aligned. More precisely, from the Two- Dimensional (2D) cross-correlation, a neighborhood of its peak value is extracted and the interpolation of both the 2D paraboloid and the two alternative 1D parabolas are computed to provide the finer misregistration estimation with a sub-pixel accuracy. The main advantage of the proposed framework is that the overall computational burden is only due to the 2D cross-correlation estimation since the parabolic interpolation is calculated with a closed-form expression. The results obtained on real recorded Unmanned Aerial Vehicle (UAV) SAR data highlight the effectiveness of the proposed approach as well as its capabilities to provide some benefits with respect to other available strategy.

AB - In this paper the problem of Synthetic Aperture Radar (SAR) images coregistration is considered. In particular, a novel algorithm aimed at achieving a fine sub-pixel coregistration accuracy is developed. The procedure is based on the parabolic interpolation of the 2D cross-correlation computed between the two SAR images to be aligned. More precisely, from the Two- Dimensional (2D) cross-correlation, a neighborhood of its peak value is extracted and the interpolation of both the 2D paraboloid and the two alternative 1D parabolas are computed to provide the finer misregistration estimation with a sub-pixel accuracy. The main advantage of the proposed framework is that the overall computational burden is only due to the 2D cross-correlation estimation since the parabolic interpolation is calculated with a closed-form expression. The results obtained on real recorded Unmanned Aerial Vehicle (UAV) SAR data highlight the effectiveness of the proposed approach as well as its capabilities to provide some benefits with respect to other available strategy.

KW - synthetic aperture radar

KW - SAR coregistration

KW - 2D cross-correlation

KW - parabolic interpolation

KW - sub-pixel coregistration

KW - UAV SAR

UR - https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=36

M3 - Article

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

SN - 0196-2892

ER -