A computationally feasible optimization approach to inverse SAR translational motion compensation

Risto Vehmas, Juha Jylhä, Minna Väilä, Jarkko Kylmälä

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

    4 Citations (Scopus)

    Abstract

    The traditional approach to inverse synthetic aperture radar translational motion compensation is to solve the problem in the two distinct parts of range alignment and autofocus. In this paper, we follow this practice and propose an approach based on the global range alignment and contrast optimization autofocus methods. The proposed range alignment procedure parametrizes the track as a spline polynomial and minimizes the loss function determined by the sum of the squared envelope differences. The necessary numerical global optimization is performed with the differential evolution algorithm. The solution of the autofocus problem is produced with first order numerical optimization, as we solve it by using an expression derived for the gradient of the loss function. In this paper, we consider the back-projection case but the proposed approach is easily extended to other reconstruction techniques. We use simulated inverse synthetic aperture radar data to demonstrate the proposed approach and to illustrate its computational efficiency.
    Original languageEnglish
    Title of host publicationProceedings of the 12th European Radar Conference (EuRAD 2015)
    PublisherIEEE
    Pages17-20
    Number of pages4
    ISBN (Print)978-2-87487-041-5
    DOIs
    Publication statusPublished - 2015
    Publication typeA4 Article in a conference publication
    EventEuropean Radar Conference -
    Duration: 1 Jan 2013 → …

    Conference

    ConferenceEuropean Radar Conference
    Period1/01/13 → …

    Publication forum classification

    • Publication forum level 0

    Fingerprint Dive into the research topics of 'A computationally feasible optimization approach to inverse SAR translational motion compensation'. Together they form a unique fingerprint.

    Cite this