Radar Target and Moving Clutter Separation Based on the Low-Rank Matrix Optimization

Research output: Contribution to journalArticleScientificpeer-review

4 Citations (Scopus)
124 Downloads (Pure)

Abstract

A novel algorithm is put forward to separate radar target and moving clutter based on a combination of the low-rank matrix optimization (LRMO) and the decision tree. Similar to the moving object detection in the field of automated video analysis, the proposed separation method, which is carried out in the range-Doppler domain, makes use of different motion variations of radar target and clutter in the spectrogram sequence. The technique is very general, but the focus of this paper is on narrowband moving clutter suppression in a weather radar. The first step in implementing this method is the generation of a range-Doppler spectrogram sequence. Then, the LRMO is applied to the obtained sequence to divide target and moving clutter into foreground and background. From the foreground sequence which is obtained by solving the LRMO, foreground frequency and spectral width are combined in a decision tree to obtain a filtering mask to mitigate the narrowband moving clutter and noise. Data collected by a polarimetric Doppler weather radar known as the IRCTR Drizzle Radar are used to validate the performance of the proposed algorithm. Moreover, its effectiveness in removing narrowband moving clutter is quantitatively assessed through comparisons with another clutter mitigation method.

Original languageEnglish
Pages (from-to)4765 - 4780
Number of pages16
JournalIEEE Transactions on Geoscience and Remote Sensing
Volume56
Issue number8
DOIs
Publication statusPublished - 2018

Keywords

  • Clutter
  • Decision tree
  • Doppler effect
  • Doppler radar
  • low-rank matrix optimization (LRMO)
  • Meteorological radar
  • narrowband moving clutter
  • Radar clutter
  • range-Doppler spectrogram sequence
  • Spectrogram
  • target and clutter separation
  • weather radar.

Fingerprint Dive into the research topics of 'Radar Target and Moving Clutter Separation Based on the Low-Rank Matrix Optimization'. Together they form a unique fingerprint.

  • Cite this