ISAR imaging of uniformly rotating targets via parametric weighted L 1 minimization

Wei Rao*, Gang Li, Xiqin Wang, Xiang Gen Xia

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

The ISAR imaging problem can be converted to a sparse recovery problem and efficiently solved by a weighted L 1 minimization algorithm, which outperforms the regular L 1 minimization substantially, provided the sparse representation is already known. However, the sparse representation of the ISAR signal is related to the target rotation rate, which is usually unknown for noncooperative target. In this paper, we propose a computationally efficient parametric weighted L 1 minimization algorithm to retrieve both the sparse representation and the ISAR image. The proposed algorithm can adaptively update the rotation rate estimation to the true value after several iterations. Numerical experiments show that the approach is efficient in estimating the target rotation rate and recovering the high resolution ISAR image.

Original languageEnglish
Title of host publication2011 3rd International Asia-Pacific Conference on Synthetic Aperture Radar, APSAR 2011
Pages363-366
Number of pages4
Publication statusPublished - 2011
Externally publishedYes
Event2011 3rd International Asia-Pacific Conference on Synthetic Aperture Radar, APSAR 2011 - Seoul, Korea, Republic of
Duration: 26 Sept 201130 Sept 2011

Publication series

Name2011 3rd International Asia-Pacific Conference on Synthetic Aperture Radar, APSAR 2011

Conference

Conference2011 3rd International Asia-Pacific Conference on Synthetic Aperture Radar, APSAR 2011
Country/TerritoryKorea, Republic of
CitySeoul
Period26/09/1130/09/11

Keywords

  • ISAR imaging
  • parametric weighted L minimization
  • rotation rate estimation
  • uniformly rotating target

Fingerprint

Dive into the research topics of 'ISAR imaging of uniformly rotating targets via parametric weighted L 1 minimization'. Together they form a unique fingerprint.

Cite this